WO1998046011A1 - Special effect apparatus and special effect method - Google Patents

Abstract

A shadow closer to reality is added to an object image with a high speed with a simple construction and simple processings. A shadow video signal V4 and a shadow key signal K4 are generated by a shadow signal generating unit (20) and the shadow key signal K4 is inputted to a real shadow signal generator (50), which generates a real shadow key by which a more realistic shadow is generated.

Description

 Specification

 Special effect device and special effect method

 The present invention relates to an image processing device, and more particularly to a special effect device that obtains a three-dimensional effect by applying an appropriate shadow to an object video signal that is three-dimensionally image-converted. Background art

 An example of a special video effect device is shown in Fig. 1. FIG. 1A is an explanatory diagram showing one of the effects of the video special effect device, and FIG. 1B is a configuration diagram of a video special effect device capable of realizing such effects.

 As shown in FIG. 1A, the video special effect device shown in FIG. 1B generates a pseudo shadow IS from an object, and combines the object 0 and the pseudo shadow IS. Here, the object is a video to be processed to apply a special effect, and the object video signal is a video signal to be processed by the special effect.

The video special effect device shown in FIG. 1B subtracts the key signal K supplied via the input terminal 101 from the background signal BG supplied via the input terminal 103, and calculates the result of this operation. A mixing circuit 102 for adding a video signal V i supplied via an input terminal 100 to the input terminal 100; a memory 104 for storing a key signal K supplied via an input terminal 101; An operation unit 105 for generating a pseudo-shadow signal IS by changing the position of the key signal K stored in the memory 104 by operating an address in the X and Y directions; From 104 the pseudo shadow signal IS is mixed with an over-verter one circuit 106 that reduces the overlapping part of the pseudo shadow signal IS and the key signal K supplied via the input terminal 101. From the mixed signal MIX from the circuit 102, the overunder circuit 1

A mixing circuit that subtracts the over-under processed output KZIS from 06, adds the color matte signal supplied via the input terminal 108 to the subtraction result, and outputs the result as the output signal OUT from the output terminal 109. It consists of 107. ,

 Next, the operation of the video special effect device shown in FIG. 1B will be described with reference to FIG. 1A.

 An object video signal V i is supplied via an input terminal 100. This corresponds to object 0 shown in FIG. 1A. The key signal K supplied via the input terminal 101 is a key signal K for keying the object video signal Vi. The mixing circuit 102 subtracts the signal corresponding to the key signal K from the background signal BG, adds the object video signal Vi to the position, and adds the mixed signal M

Generate 1 X.

 On the other hand, the key signal K supplied through the input terminal 101 is supplied to the memory 104.

 When the addresses in the X and Y directions are changed by the operation unit 105, the operation unit 105 changes the position of the key signal K in the memory 104. The key signal K is read from the memory 104 and is supplied to the over-under circuit 106 as a pseudo shadow signal IS. Here, the screen of the pseudo shadow signal IS is an image in which the pseudo shadow I shown in FIG. 1A is not hidden by the object 0. The keyer signal K and the pseudo-shade signal IS are supplied to the inverter circuit 106. The overunder circuit 106 subtracts only the overlapping portion of the key signal and the pseudo shadow signal IS from the pseudo shadow signal IS. Over-under circuit 1 0

The overhander processing output KZIS from 6 corresponds to the pseudo shadow IS shown in FIG. 1A.

Over-Znder processing output KZIS and mixed signal MIX are Each is supplied to the mixing circuit 107. In the mixing circuit 107, the over-Z under-processing output KZIS from the over-Z under-circuit 106 is subtracted from the mixed signal MIX from the mixing circuit 102, and the input terminal 10 The power cutout signal supplied via 8 is added, and the addition result is output from the output terminal 109 as an output signal 〇UT. This output signal is the entire image shown in FIG. 1A.

 By the way, the pseudo shadow I shown in Fig. 1A is simply a shift of the key signal generated from the object 〇. That is, the pseudo-shaded image shown in FIG. 1A is a two-dimensional image. Therefore, in order to get closer to the actual shadow, it is necessary to use a three-dimensional image. In this case, it is necessary to generate a rotated shadow for the object and combine the shadow with the object. For this purpose, it is necessary to prepare one video special effect device for the object and one video special effect device for the shadow, and change the rotation axes of the object and the shadow before combining. In this case, the required operation technology is very advanced because it is the synthesis of images generated by different video special effect devices.

Furthermore, the longer the distance of the real shadow from the object is, the blurred the outline of the shadow becomes and the lighter the color of the shadow becomes. Therefore, in order to satisfy the desire to add a shadow similar to a real shadow by a special effect, processing corresponding to the distance from the object must be performed. However, at present, it is extremely difficult to perform such processing easily with a special video effect device. Of course, it is possible to generate an image of a shadow that is very close to the real shadow by using the technology of 3-computer graphics. However, computer graphics take an enormous amount of time to generate a single image. Therefore, it can be released using computer graphics. It is almost impossible to process images that are input without interruption, such as transmission and briefing. Disclosure of the invention

 A special effect device according to the present invention is a special effect device for performing a special effect process on an object indicated by a video signal and a shadow on the object, wherein the gain control means controls a gain of an image of a shadow of the object; Filtering means for filtering the image, control means for controlling the gain of the gain control means in accordance with the shadow depth information, and controlling the filtering characteristics of the filtering means in accordance with the shadow depth information; and Means for synthesizing an image of the shadow of the object controlled and output by the means, an image of the object, and an image serving as a background of the object.

 In the special effect device according to the present invention, the control means has a storage means for storing gain characteristic data according to the depth information and filtering characteristic data, and the control means is stored in the storage means. The gain of the gain control means is controlled in accordance with the gain characteristic data, and the filtering characteristic of the filtering means is controlled in accordance with the filtering characteristic data stored in the storage means.

 In the special effect device according to the present invention, the filtering means includes a low-pass filter.

A special effect device according to the present invention is a special effect device for performing a special effect process on an input source video signal. The special effect device performs a first image conversion process on the source video signal to indicate an object image. Object signal generating means for generating a shadow signal; and shadow signal generating means for performing a second image conversion process on the source video signal to generate a shadow signal corresponding to the target image. An object signal output from the object signal generating means; The shadow signal output from the shadow video signal generation unit is input, and the object signal, the shadow signal, and the background signal corresponding to the source video signal are combined and output. Synthesizing means for outputting an output video signal, wherein the shadow signal generating means includes: gain control means for controlling a gain of the shadow and dough signals corresponding to the source video signal; and the shadow signal , A depth information generating means for generating depth information corresponding to the shadow signal, and a gain of the gain control means based on the depth information from the depth information generating means. And a control stage for controlling the filtering characteristics of the filtering means based on the depth information.

 In the special effect device according to the present invention, the shadow signal input to the shadow signal generating means is output to the gain control means as a shadow signal whose gain is controlled, and the gain is controlled. The shadow signal thus input is input to the filtering means, is filtered, and is output from the filtering means as a shadow signal.

 In the special effect device according to the present invention, the filtering means includes a low-pass filter.

 In the special effect device according to the present invention, the control means has storage means for storing gain characteristic data and filtering characteristic data corresponding to the depth information from the depth information generating means, The filtering characteristic data is supplied from the storage means to the gain control means and the filtering means in accordance with the depth information, respectively.

In the special effect device according to the present invention, the object signal generating means may include a storage means for storing the input source video signal, and the source video signal stored in the storage means may be read out in a predetermined unit. Read address generating means for generating a read address, wherein the first image conversion process is performed by the read address. The object video signal read from the memory means is processed as a target image.

 In the special effect device according to the present invention, the shadow video signal generation means includes a storage means for storing the input source video signal, and a readout unit for reading the source video signal stored in the storage means in a predetermined unit. A read address generating means for generating a dress, and wherein the second image conversion processing is such that the shadow signal read from the memory means by the read address is used as the target image. Is processed to be output as a shadow signal corresponding to.

 In the special effect device according to the present invention, the synthesizing unit may be configured to receive the object signal from the object signal generating unit and the shadow signal from the shadow signal generating unit, and to input each of the signals. And a second synthesizing unit that receives the mixing unit and the background signal and outputs an output video signal. .

The special effects device according to the present invention is a special effects device for performing special effects processing on an input source video signal and a source key signal corresponding to the source video signal. , An object signal generating means for performing a first image conversion process to generate an object video signal and an object key signal indicating a target image, and the source video signal and the source key signal are input. Then, a shadow signal generating means for performing a second image conversion process to generate a shadow video signal and a real shadow signal corresponding to the target image, and an object output from the object signal generating means The video signal and the object signal are input and the shadow signal generating means The shadow video signal and the real shadow signal Synthesizing means for inputting a background signal corresponding to the source video signal, synthesizing the input signals, generating an output video signal, and outputting the output video signal. Wherein the shadow signal generating means includes: a shadow key signal generating means for generating a shadow key signal corresponding to the target image with respect to the input so-called ski key signal; and the shadow key signal generating means. Gain control means for controlling the gain of the shadow key signal output from the means, filtering means for filtering the shadow key signal, depth information generating means for generating depth information corresponding to the shadow signal, The gain of the gain control means is controlled based on the depth information from the depth information generating means, and the filtering characteristic of the filtering means is controlled based on the depth information. And control means for performing the operation.

 In the special effect device according to the present invention, the filtering means includes a low-pass filter.

 In the special effect device according to the present invention, the shadow key signal output from the shadow key signal generation means of the shadow signal generation means is output as a shadow key signal whose gain is controlled by the gain control means, and The controlled shadow key signal is inputted to the above-mentioned filtering means, and the filtering characteristic is controlled and outputted as a real shadow signal.

 In the special effect device according to the present invention, the control means of the shadow signal generator has storage means for storing gain characteristic data and filtering characteristic data corresponding to the depth information from the depth information generating means. Then, the gain characteristic data and the filtering characteristic data are supplied to the gain control means and the filtering means, respectively, in accordance with the depth information.

In the special effect device according to the present invention, the synthesizing means may include the object video signal from the object signal generating means and the object video signal. In addition to the input of the cutout signal, the shadow video signal from the shadow signal generating means and the real shadow signal are input, and the input signals are combined to produce a mixed video signal. First synthesizing means for generating the mixed video signal, the mixed key signal and the background signal from the first synthesizing means, and generating the output video signal. And 2 combining means.

 In the special effect device according to the present invention, the object signal generating means may include first memory means for storing the source video signal and second memory means for storing the source key signal. The source video signal and the source key signal stored in the first memory means and the second memory means, respectively, from the respective memory means. Read address generating means for generating the read address to supply the read address to the first memory means and the second memory means. In the image conversion processing, the object video signal and the object key signal read from the first memory means and the second memory means, respectively, by the read address are used for the above purpose. Appear as an image It is intended to be.

In the special effect device of the present invention, the shadow signal generating means may include a third storage means for storing the source video signal and a fourth memory means for storing the source key signal. And a read address for reading the source video signal and the source key signal stored in the third memory means and the fourth memory means, respectively, from the respective memory means. And a read address generating means for supplying the read address to the third memory means and the fourth memory means, and the second image conversion processing comprises: The third memory means and the second memory means The above-mentioned shadow video signal and the above-mentioned shadow key signal read out from the memory means 4 are output, and the above-mentioned shadow video signal is output as a shadow video signal corresponding to the above-mentioned target image. is there.

 The special effect method according to the present invention is a special effect method for performing a special effect process on an object formed by an input video signal and a shadow of the object, wherein a gain step for controlling a gain of an image showing the shadow of the object is provided. A filtering step of filtering the shadow image; controlling the gain of the gain step in accordance with the shadow depth information; and adjusting a filtering characteristic of the filtering step in accordance with the shadow depth information. A control step of controlling; combining the shadow image of the object, which is controlled and output by the control step, the image of the object, and the background image of the object to output an output video signal It consists of steps.

 In the special effect method according to the present invention, in the control step, the gain is controlled in the gain step by gain characteristic data corresponding to the depth information stored in a storage unit, and the depth stored in the storage unit. In the filtering step, the shadow image is filtered by the filtering characteristic data corresponding to the information.

The special effect method according to the present invention is a special effect method for performing a special effect process on an input source video signal, wherein the first image conversion process is performed on the source video signal to indicate a target image. An object signal generating step of generating an object signal; and a shadow signal generating step of performing a second image conversion process on the source video signal to generate a shadow signal corresponding to the target image. And the object signal, the shadow signal, and the background signal corresponding to the source video signal, and combine the input signals to output an output video signal. The shadow signal step includes a gain control step of controlling a gain of the shadow signal corresponding to the source video signal, and a filtering step of filtering the shadow signal. A depth information generating step for generating depth information corresponding to the shadow signal, and controlling the gain of the gain control step based on the depth information from the depth information generating step. And a control step of controlling the filtering characteristics of the filtering step based on the depth information.

 The special effect method according to the present invention includes an object signal generation step of performing a first image conversion process on an input source video signal to generate an object signal indicating a target image; A shadow signal generation step of performing a second image conversion process to generate a shadow signal corresponding to the target image, an object signal generated in the object signal generation step, and the shadow video signal The shadow signal generated in the generation step is input, and the object signal, the shadow signal, and the background signal corresponding to the source video signal are combined to output an output video signal. A synthesizing step, wherein the shadow ゥ signal generating step includes the shadow corresponding to the source video signal. A gain control step for controlling the gain of the signal, a filtering step for filtering the shadow signal, a depth information generation step for generating depth information corresponding to the shadow signal, and a depth information generation step. A control step of controlling the gain of the gain control step based on the depth information from the step and controlling the filtering characteristic of the filtering step based on the depth information. The above-mentioned filtering step is performed by a one-pass filter.

In the special effect method according to the present invention, the control step may include: A step of storing gain characteristic data and filtering characteristic data according to the depth information from the information generating step, wherein the gain characteristic data and the filtering characteristic data are stored in accordance with the depth information from the storing step. These are supplied to the gain control step and the filtering step, respectively.

 In the special effect method according to the present invention, the object signal generating step includes: a storage step of storing the input source video signal; and the source video signal stored in the storage step in a predetermined unit. A read address generating step for generating a read address so that the read address can be read out, and the first image conversion processing is performed by the object read out from the storage step by the read address. The video signal is processed as a target image.

 In the special effect method according to the present invention, the shadow video signal generation step includes: a storage step of storing the input source video signal; and the source video signal stored in the storage step. A read address generation step for generating a read address to be read in a predetermined unit, wherein the second image conversion processing is performed by the memory read out from the memory means by the read address. The video signal is processed so as to be output as a shadow signal corresponding to the target image.

 In the special effect method according to the present invention, in the synthesizing step, the object signal from the smart signal generating step and the shadow signal from the shadow signal generating step are input and input. A first combining step of combining the above signals to output a mixed signal, and a second combining step of receiving the mixed signal and the background signal and outputting an output video signal. is there.

The special effect method according to the present invention comprises the steps of: In a special effect method in which a special effect process is performed on a source key signal corresponding to a source video signal, the source video signal and the source key signal are input, a first image conversion process is performed, and a target image is processed. A source signal generation step for generating an object video signal and an object key signal to be shown and the source, the video signal, and the source key signal are input, and a second image conversion process is performed. A shadow video signal generating step for generating a shadow video signal and a real shadow key signal corresponding to the target image, and an object video signal and an object key signal output from the object signal generating step are input. Together with the shadow video signal from the shadow signal generation step and the real shadow A key signal and a background signal corresponding to the source video signal are input, and the input signals are synthesized to generate and output an output video signal; and A shadow key signal generating step of generating a shadow key signal corresponding to the target image with respect to the input source key signal; A gain control step for controlling the gain of the shadow key signal output from the shadow key signal generation step; a filtering step for filtering the shadow key signal; and generating depth information corresponding to the shadow signal. A depth information generating step; and controlling the gain of the gain control step based on the depth information from the depth information generating step. And a control step of controlling a filtering characteristic of the filtering step based on the depth information.

 In the special effect method according to the present invention, the filtering step is performed by a one-pass filter.

The special effect method according to the present invention includes a shadow key output from the shadow key signal generation step in the shadow signal generation step. The signal is output as a shadow key signal whose gain has been controlled in the gain control step, and the shadow key signal whose gain has been controlled is input to the filtering step where the filtering characteristic is controlled and the real shadow key signal is output. It is output as a signal.

 In the special effect method according to the present invention, in the shadow signal generation step and the step, the gain control data and the filtering characteristic data corresponding to the depth information from the depth information generation step are stored. A special effect method comprising a storage step, wherein the gain characteristic data and the filtering characteristic data are supplied to the gain control step and the filtering step, respectively, according to the depth information.

 In the special effect method according to the present invention, in the synthesizing step, the object video signal and the object key signal from the object signal generating step are input and the shadow signal generating step is performed. A first synthesizing step of receiving the shadow video signal and the real shadow key signal and synthesizing the input signals to generate a mixed video signal and a mixed key signal; and the first synthesizing step. And a second synthesizing step of receiving the mixed video signal, the mixed key signal, and the background signal from each other to generate the output video signal.

In the special effect method according to the present invention, the object signal generating step includes a first memory step for storing the source video signal and a second memory step for storing the source key signal. And the source video signal and the source key signal stored in the first memory step and the second memory step, respectively, from the respective memory steps. A read address generating step of generating a read address for reading and supplying the read address to the first memory step and the second memory step. The first image conversion process is performed by the readout key. The object video signal and the object signal read from the first memory step and the second memory step, respectively, are output as the target image by the address. It is something.

 In the special effect method according to the present invention, in the shadow signal generating step, a third storage step for storing the source video signal and a fourth memory step for storing the source key signal are included. The source video signal and the source key signal stored in the third memory step and the fourth memory step, respectively, from the respective memory steps. A read address generating step for generating the read address to supply the read address to the third memory step and the fourth memory step; In the image conversion process, the shadow video signal and the shadow key read from the third memory step and the fourth memory step by the read address, respectively, are used. A signal is output and the shadow video signal is output as a shadow video signal corresponding to the target image.

The present invention provides a shadow generating means for generating a shadow image, a synthesizing means for synthesizing an object image to which a shadow is to be applied, the above-mentioned shadow image, and a background image. Gain characteristic data; storage means for storing filtering characteristic data; gain control means for controlling the gain of the shadow image; filtering means for filtering the shadow image; and the shadow image and the object The ratio xZy between the distance X from the image of the object and the distance y between the virtual light source and the image of the object is determined, and the gain control method is performed according to the gain characteristic according to the ratio XZy. A control means for controlling the gain in the stage and controlling the filtering characteristics of the filtering means in accordance with the filtering characteristics in accordance with the ratio xZy. And a step.

 The gain in the gain control means is controlled in accordance with the gain characteristic in accordance with the ratio XZy, and the filtering characteristic in the filtering means is controlled in accordance with the filtering characteristic in accordance with the ratio xZy. An image of a close shadow is generated, and the shadow image, the object image, and the background image are combined. BRIEF DESCRIPTION OF THE FIGURES

 FIG. 1 is a diagram for explaining the background art.

 FIG. 2 is a block diagram showing the entire configuration of the special effect device of the present invention.

 FIG. 3 is a diagram for explaining a world coordinate system defined in the special effect device of the present invention.

 FIG. 4 is a diagram for explaining a conversion process for obtaining an object video signal.

 FIG. 5 is a diagram showing the relationship between the address on the frame memory corresponding to the object video signal and the address on the monitor screen.

 FIG. 6 is a diagram showing an array of pixel data on the monitor screen and in the frame memory for indicating the meaning of the parameter Η.

 Figure 2 is a diagram showing the address space of the frame memory when the perspective method is used.

 FIG. 8 is a conceptual diagram showing the relationship between the world coordinate system and the shadow coordinate system.

The figure shows the conversion process for obtaining a shadow video signal in the point light source mode, and the address and monitor screen on the frame memory corresponding to the shadow video signal. FIG. 6 is a diagram showing a relationship with an address on a plane. FIG. 10 is a diagram for explaining a perspective transformation process for obtaining a three-dimensional shadow video signal from a three-dimensional object video signal in the point light source mode.

 FIG. 11 is a flowchart showing a procedure of a conversion process for obtaining a point light source and a three-dimensional object video signal.

 FIG. 12 is a diagram for explaining a conversion process for obtaining a shadow video signal in the parallel light source mode.

 FIG. 13 is a flowchart showing a procedure of a conversion process for a parallel light source and a three-dimensional object video signal.

 FIG. 14 is a diagram for illustrating a relationship between spherical coordinates and rectangular coordinates in a world coordinate system and a shadow coordinate system, respectively.

 FIG. 15 is a diagram for explaining a perspective transformation process for obtaining a three-dimensional shadow video signal from a three-dimensional object video signal in the parallel light source mode.

 FIG. 16 is a diagram for explaining a perspective transformation process for obtaining a virtual three-dimensional shadow video signal from a virtual three-dimensional object video signal.

 FIG. 17 is a diagram for explaining an origin setting mode for automatically setting the origin of the shadow coordinate system.

 FIG. 18 is a diagram showing a configuration of a real shadow generator.

 FIG. 19 is a diagram showing a target image and a diagram showing the relationship between parameter H and gain and filtering characteristics.

 FIG. 20 is a diagram for explaining the combining process of two intersecting images.

 FIG. 21 is a block diagram showing the detailed configuration of the combiner. FIG. 22 is a block diagram showing a detailed configuration of the priority signal generating circuit of FIG. 21.

Figure 23 illustrates the operation of the priority signal generator of Figure 22. FIG. 4 is a signal waveform diagram provided for FIG. BEST MODE FOR CARRYING OUT THE INVENTION

 (1) Overall configuration

 First, the configuration of the special effect device 1 of the present invention will be described with reference to FIG.

The CUP 8 is a processor for controlling all the circuits of the special effect device 1. The CPU 8 receives each parameter obtained by operating the control panel 5 by the operator via the interface circuit (IZF) 5a and the data bus, and based on this parameter. To control each circuit. From this control panel 5, the perspective value P _z , X in the shadow coordinate system

S-axis, Y- _S- axis and Z- _S- axis rotation angles (<9 _X , θ V, 、, origin of shadow coordinate system (x _so , y so, z so)) types of light sources indicating how the position of the point light source (x _L, y zeta Mr), the position of the collimated light source (r, alpha, beta), Pas lame Isseki ru ~ r ₃₃ on 3D conversion, 1 X, 1 γ, 1 _z, and s are input, and the parameters will be described later.The CPU 8 receives these parameters input from the control panel 5 and receives these parameters. The parameters are read out in real time and reflected in the calculation of the address.Specifically, the CPU 8 monitors the change of the parameter supplied from the control panel 5 at a frame cycle. And parameters for calculating the read address based on the supplied parameters (! ^ To ゎ; ^, b, _{~ B 33 ', b, ~} b 33 ") frame period is not calculated. Ru a. Therefore, varying these parameters to re Alta Lee beam at full rate A period in response to operation of the operator And apply real-time special effects to the source video signal according to the changed parameters. it can.

 Note that the special effect device of the present invention allows a desired light source to be selected from a point light source and a parallel light source by inputting the type of the light source from the control panel 5. In the following description, the mode in which an object shadow is generated by a point light source is referred to as a point light source mode. I will call it.

 Further, the CPU 8 controls each circuit and calculates a read address based on a program stored in a ROM (Read Only Memory) 6 provided as a program memory. Also

Similarly, based on data stored in a RAM (Random Access Memory) 7 provided as work memory, each circuit is controlled and a read address is calculated.

 The object signal generator 10 receives the source video signal V 0 from the outside, and receives the source video signal V 0. Is converted to 3D

, To produce a two-dimensional object video signal V _3. Also, the object signal generator 10 outputs the source video signal V. It receives the source key signal K _D for keying, and receives the source key signal K in the same way as the source video signal. Is subjected to three-dimensional conversion processing to generate an object key signal K ₂ . Specifically, the object signal generation unit 10 generates the source video signal V. Frame memory 12 for temporarily storing data, and a soft key signal キ for keying this source video signal. Frame memory 13 for temporarily storing data, and read addresses calculated for frame memory 12 and frame memory 13 corresponding to the three-dimensional conversion operation. (

, _M , Y _M ).

Frame memory 12 is the supplied source video signal V. One It is a memory for storing it temporarily. The frame memory 12 is supplied with a sequential write address from a write address generation circuit (not shown), so that the supplied source video signal V is supplied to the frame memory 12. Is stored as it is without deformation. In addition, a read address (Χ _,, Υ _Μ ) calculated in accordance with the three-dimensional conversion operation is supplied from the read and address generation circuit 14 to the frame memory 12. Thus, the object video signal V _{2 that} has been subjected to the three-dimensional conversion processing for each frame is output from the frame memory 12. The output object video signal V ₂ is sent to the mixer 30.

Frame memory 13 is a source key signal K for keying the source video signal V 0. This is a memory for temporarily storing data. In this frame memory 13, the same write address as the sequential write address supplied to the frame memory 12 is supplied. Similarly to this, the frame memory 13 receives the supplied sousky signal Κ. Is stored as it is without deformation. In addition, the full Remume mode Li 1 3, frame Note Li 1 2 is supplied to the same address as the read address (Χ _Μ, Υ Μ) so is supplied from the full Remume mode Li 1 3 3D transformed object Toki first signal kappa ₂ the same as converted 3-dimensional and object Tobideo signal V ₂ which is output. The outputted object signal # ₂ is sent to the combiner 30.

Read § address generating circuit 1 4, supplied from the scan click rie N'a dress generating circuits 9 in Sequence emission sheet catcher Le. To the model two tasks Lean add-less (X _S, Y _S) on at 3 and , based on the c ° lame Ichita b ~ b _{3 3} of by the CPU 8 connexion computed image transformation matrix, the read add-supplied to the full Remume mode Li 1 2 and full Remume mode Li 1 3 Generate the address (Χ Μ, Y _M ). This read address generation circuit 14 The specific calculation in will be described later.

 The above is the configuration of the object signal generation unit 10. Next, the configuration of the shadow signal generation unit 20 will be described.

The shadow signal generation unit 20 is a circuit for generating a shadow video signal and a shadow signal. First, when the point light source mode is selected, the shadow signal generator 20 outputs the source video signal V supplied from the outside. The receipt, by three-dimensional conversion processing Matsudo processed source Subideo signal in the shadow of the color, to generate a shut Doubideo signal V _4. Further, the shadow signal generator 20 generates a source key signal K for keying the source video signal. The receive Ri, just as with three-dimensional conversion processing Shah Doubideo signal V _4, to generate a shut synchronizing first signal kappa _4. When the parallel light source mode is selected, the shadow signal generator 20 outputs the source video signal V supplied from the outside. And the source video signal matted to the color of the shadow is subjected to three-dimensional conversion processing to generate a shadow video signal V 6. Further, the shadow signal generation unit 20 generates a source key signal for keying the source video signal. And performs a three-dimensional conversion process in the same manner as the shadow video signal V ₆ to generate a shadow key signal K s. More specifically, the shadow signal generation section 20 has a circuit configuration similar to that of the object signal generation section 10 and includes the source video signal V. A color matte generation circuit 21 for matte processing, a frame memory 22 for temporarily storing a matted source video signal, and a source key signal K. Memory 23 for temporarily storing data, and reading for supplying the calculated read address to the frame memory 22 and the frame memory 23. The color matte generation circuit 21 has an address generation circuit 24 and a source video signal V0. By processing the source video signal V. This is a circuit to make the color of the shadow into a shadow. In the simplest case, the source video signal

V. By lowering the saturation and luminance levels of the source video signal, the color of the source video signal approaches that of the shadow (black).

 The frame memory 22 is a memory for temporarily storing the source video signal and the matte-processed signal. Since a sequential write address from a write address generator (not shown) is supplied to the frame memory 22, the matte-processed software is supplied to the frame memory 22. Source video signal is stored as it is without image deformation. In the point light source mode, this frame memory

The 2 reads address generating circuit 2 4 from the three-dimensional conversion operation及beauty point light source is calculated on the basis of the read-out add-less (X _M ', YM') because is supplied, the frame memory 2 2 from, Shah Doubideo signal V ₄ which has been converted 3D is output. In the parallel light source mode, the read address (2 _Μ ，, Υ _Μ ) calculated from the read address generation circuit 24 based on the three-dimensional conversion operation and the parallel light source is stored in the frame memory 22. Since this is supplied, the frame memory 22 outputs a three-dimensionally converted shadow video signal V 6.

Frame memory 2'3 is the source key signal K for keying the source video signal Vn. This is a memory for temporarily storing data. Since the same address as the sequential write address supplied to the frame memory 22 is supplied to the frame memory 23, the supplied address is supplied to the frame memory 23. Sawski — signal. Are stored as they are without image conversion. In the point light source mode, the frame memory 23 is supplied with the same address (XM ', Υ _供給 ') as the read address supplied to the frame memory 22. From frame memory 23, three-dimensional conversion is performed. In the same manner as the shadow video signal V ₄ , a three-dimensionally converted shadow key signal K ₄ is output. In the parallel light source mode, the same address (Χ _Μ ，, Υ _Μと) as the read address supplied to the frame memory 22 is supplied to the frame memory 23. The frame memory 23 outputs a three-dimensionally converted shadow key signal K _{6 in the} same manner as the three-dimensionally converted shadow video signal V ₆ .

The read address generation circuit 24 is a circuit for generating a read address to be applied to the frame memory 22 and the frame memory 23. When a point light source mode, scan click rie N'a monitor scan from de-less onset raw circuit 9 is supplied to the sequencer Nsha Torque rie down 3 on at the § de-less (X _s, Y _s) and, in the CPU 8 parameter b of Yotsute computed images transformation matrix, 'based on the read add-less _{(X M' ~ b 33,} YM ') to generate a. During directional light mode one de, add-less (X _S, Ys) of the on the monitor scan click rie down 3 from the scan click rie N'a de-less generation circuit 9 Ru is supplied to the sequencer Nsha Le and, in CPU 8 A read address (X _M 〃, YM ″) is generated based on the image conversion matrix parameters b, to b ₃₃ 〃 calculated in the above manner. Typical operations will be described later.

The real shadow signal generator 50 receives a shadow key signal (K ₄ in the point light source mode and Κ _{6 in} the parallel light source mode) output from the frame memory 23, and receives the input shadow signal. outputs a; (when source mode is kappa ₄ ', directional light mode kappa _6' point ^) against the dough signal, more realistic shadow Riarusha Douki signal for generating a. Details of the real shadow signal generator 50 will be described later.

 The above is the configuration of the shadow video signal generation unit 20 ′.

The screen address generation circuit 9 is used for the monitor screen 3 screen. This is a circuit for dressing the entire lean surface in the order corresponding to the raster scan order. Specifically, a screen address (X s, Y s) is generated based on the internally generated horizontal synchronization signal and vertical synchronization signal.

The compiler 30 is a circuit for mixing signals supplied from the object signal generator 10 and the shadow signal generator 20. In the point light source mode, the compiler 30 receives the object video signal V ₃ and the object key signal K ₂ output from the object signal generator 10, and outputs the object video signal V ₂ and the object key signal K ₂ from the shadow signal generator 20. Receives the shadow video signal V ₄ and the real shadow signal K ′, and mixes the objet video signal V ₂ and the shadow video signal V ₄ with the mixed video signal V _MIX ′ and the observ video signal K ₂ and Li Arusha synchronizing first signal kappa ₄ - produces the 'and the mixture was mixed keys signal kappa _{Micromax iota chi'.} Specifically, mixed bidet O signal V _MIX 'and mixed key signal K _{MI X',} the following equation

V „ _IX '= K ₂ V ₂ + (1-K ₂ ) Κ' V

 K MIX '= 1-(1-Κ 2) (1 one Κ'

It can be expressed as. In the parallel light source mode, the object video signal V ₂ and the object key signal # ₂ output from the object signal generator 10 are received, and the shadow video signal output from the shadow signal generator ₂₀ is received. It receives signals V and Li Arushi catcher Douki signal K '6, Obujiweku preparative bidet and old signals V 2 and Shah Doubideo signal V ₆ and the mixture was mixed video signal V _mIX ", Obujiweku Toki first signal K ₂ and Li Arusha Douki Generate a mixed key signal K _{MI X} ^ mixed with the signal K '6.

, The mixed video signal V _MIX 〃 and the mixed key signal K _MIX "are expressed by the following equation: V′−K ₂ V ₂ + (1−K ₂ ) K ^/ ₆ V ₆

K "= 1-(1-K 2) (1 — K '6) ·' · · (b) It can be expressed as. Mixer 4 which will be described later in detail co Nbaina 3 0 0 is a circuit for mixing the supplied Bakkugurau down Dobideo signal V _BK from the mixed video signal and mixed key signal and an external output from the co Nbaina 3 0 is there. In the point light source mode, it receives the mixed video signal V _MIX ′ and the mixed key signal K _MIX ′ output from the mixer 30 and receives the externally supplied background video signal V _BK . The mixed video signal V _MIX ′ and the background video signal V _BK are mixed based on the mixed key signal K _MIX ′ to generate an _output video signal V _0UT ′. More specifically, this output video signal VOUT 'is given by the following equation:

V 0 UT = KMIX VMIX + (1 — MIX) \ BK · · · · · (C) In the point light source mode, while receiving the mixed video signal V _MIX "and the mixed key signal K _MIX 'output from the mixer 30, the background video signal V

_BK , and mixes the mixed video signal V _MIX "and the background video signal V _BK based on the mixed key signal K _{MIX 、} to generate an _output video signal V _0UT ". Specifically, this output video signal V. _UT 'is

VOUT "-KMIX 'VM ix" + (1-KMIX ") VBK' · · · (d) Generated output video signal V _ουτ

'Or V. _UT 'is output to the outside and displayed on monitor screen 3.

 (2) Definition of World Coordinate System

First, the world coordinate system used in the description of the present invention will be described with reference to FIG. This world coordinate system is a three-dimensional rectangular coordinate system composed of the X, Υ, and Ζ axes. In other words, as shown in Fig. 3, the Υ plane defined by the X axis and the Υ axis orthogonal thereto Assuming that the screen surface 3 exists on the surface, the X axis is defined as the horizontal (left / right) direction of the screen surface 3, and the Y axis is vertical (up / down) of the screen surface 3. ) Direction is defined.

 In addition, the depth direction of the screen surface 3 is defined as the positive direction of the Z axis orthogonal to the XY plane, and the front side of the screen surface 3, that is, the screen surface is viewed. The side where the viewpoint PZ exists is defined as the negative direction of the Z axis.

 Further, it is defined that the center of the screen plane 3 matches the origin of the world coordinate system including the X axis, the Y axis, and the Z axis.

 Virtual coordinate values are set on the X-axis that are continuous from the inside (origin) of the screen area to the left and right sides, and the X-axis in the screen area has the viewpoint PZ The virtual coordinate values between "1 4" and "+4" are set from left to right when viewing the screen surface 3 from.

 In addition, virtual coordinate values that are continuous from the inside (origin) of the screen area in the vertical and external directions are set on the Y axis, and the Y axis in the screen area is provided with a viewpoint. Virtual coordinates between “13” and “+3” are set from below to above when viewing the screen surface 3 from PZ.

 Further, the operator's viewpoint position P Z is virtually set on the Z axis at a position where its coordinate value is “1 16”.

() Description of Conversion Processing for Generating Object Video Signal First, the conversion processing for generating an object video signal V ₂ from a source video signal V □ will be described with reference to FIGS. 4A and 4B. First, the source video signal V, which is two-dimensional data. Is stored in the frame memory 12 as it is without image conversion. Therefore, as shown in FIGS. 4A and 4B, the source video signal V. Is on the XY plane of the world coordinate system, Source video signal V on screen 8 present in the Y plane. Image is displayed.

 Incidentally, Fig. 4 (a) shows a state where the XY plane is viewed from the viewpoint PZ on the Z axis in the space represented by the three-dimensional coordinates of the world coordinate system. The displayed image appears. FIG. 4B shows a state where the XZ plane is viewed from the viewpoint position on the positive side of the Y axis in a space represented by three-dimensional coordinates in the world coordinate system. Therefore, the source video signal V 0 existing on the XY plane overlaps the screen surface 3.

 Such source video signal V. In response, the operator operates the controls on the control panel to perform 3D image conversion processing in the world coordinate space. That is, a three-dimensional conversion matrix T consisting of parameters set for each frame. The source video signal V by operating the operator. The source video signal V is applied to each pixel. Is converted to a three-dimensional spatial position. In FIG. 4, the video signal subjected to the three-dimensional image conversion processing is represented as a three-dimensional object video signal V ,. The source video signal V is used for the three-dimensional conversion in the cases of Figs. This is an example in which is rotated about 45 ° about the X-axis as the center of rotation, and further translated in the positive direction of the Z-axis.

The three-dimensional transformation matrix T _Q used for three-dimensional transformation is

Γ 1! Γ 1 2 0

 Γ 2 1 Γ 2 2 2 3 0

 T o = (1)

 r 3 1 I 3 2 0

It is expressed as 1 X 1 Y 1 s. This 3D transformation matrix Τ. The conversion parameters,, Γ ^ used for the source video signal V. Around the X axis , Element for rotating around the Y axis and around the z axis, the source video signal V. For scaling the scale in the X-, Y-, and Z-axis directions, respectively, and the source video signal V. A parameter including elements for skewing in the X-axis direction, the Y-axis direction, and the z-axis direction, respectively. Lame data 1 _x , 1 _Y ,, 1 _z are

The source video signal V is a parameter including elements for translating the source video signal in the X-axis, Y-axis, and Z-axis directions, respectively. This is a parameter that includes elements for uniformly enlarging and reducing the size of the whole in each of the three-dimensional axial directions.

 Note that this transformation matrix T. Is a 4-by-4 matrix because the coordinate system for rotation transformation and the coordinate system for translation transformation and scaling transformation are represented in the same single coordinate system. It is called the next coordinate system (Homogeneous Color or dmate).

Thus, the source video signal V _Q existing in the XY plane is transformed into the three-dimensional conversion matrix Τ. Is converted to the three-dimensional position represented by the three-dimensional object video signal V, by the following perspective transformation processing.

This perspective transformation processing is to transform the three-dimensional object video signal V, with the perspective transformation matrix P, as shown in FIGS. Is a transformation process that allows the XY plane to be seen through. In other words, when the 3D object video signal V, is viewed from the virtual viewpoint PZ on the Z axis, the conversion is to obtain the image of the video signal seen through the XY plane. In Figure 4 B, represents thus the perspective video signal to the XY plane, the 2-dimensional and the object video signal V _2. In the case of FIG. 4B, the object video signal V ₂ seen through the screen plane 3 on the XY plane is a three-dimensional object on the depth side of the screen plane 3 as viewed from the virtual viewpoint PZ. This is a video image as if the video signal V, exists. The perspective transformation matrix P of. Is:

 1 0 0 0

 0 1 0 0

 P (2)

 0 0 0 P z

 0 0 0 1

Is represented by This perspective transformation matrix P. 'S parameter P _z, when seen through the three-dimensional conversion video signal V ₂ on the XY plane is a perspective Bae Kuti blanking value for applying the perspective. That is, in the case of FIG. 4B, the three-dimensional object video signal V, in the three-dimensional space, is inclined by about 45 ° with respect to the XY plane, and when viewed from the virtual viewpoint PZ, The part that is far away looks small, and the part that is close looks large. Thus by using the perspectives blanking the shown to para menu over data P _z, Obujiweku preparative video signal V ₂ of the two-dimensional, which is transparent to the XY plane, three-dimensional Obujiweku preparative video signal V, and the virtual viewpoint in the three-dimensional space It is converted according to the distance from PZ.

The transformation position of the three-dimensional object video signal V on the screen plane 3 by the perspective transformation is determined by the distance between the virtual viewpoint PZ and the screen plane 3, and the virtual viewpoint PZ and the three-dimensional object. It changes according to the distance from the virtual video signal V, and the perspective conversion according to the position of the virtual viewpoint PZ is performed by the operator setting the perspective value _Pz according to the position of the virtual viewpoint PZ. It can be carried out. Normally, since the position of the viewpoint PZ is the coordinate value “1 16” on the z-axis, the passive value P _z is set so that “1/16” is a standard value.

As described above, the source video signal V. Conversion process to generate a two-dimensional Obujeku preparative video signal V ₂ from the three-dimensional Obujiweku protruded by a three-dimensional transformation matrix T o from the source Subideo signal VQ A spatial image conversion step for obtaining the three-dimensional object video signal V, obtained by the spatial image conversion step. By composed of the perspective conversion step of converting two-dimensional in Obujiweku Tobideo signal V _2. Therefore, the source video signal V. The transformation matrix Tobj for obtaining a two-dimensional object video signal V a from the matrix is a three-dimensional transformation matrix T _Q and a perspective transformation matrix P. As a multiplication expression with

 Tobj = T 0P 0

Is represented by

Here, the image processing apparatus using the special effect device of the present invention writes the two-dimensional source video signal VQ supplied from the outside to the frame memory 12 once, and By supplying the two-dimensional read address (X _s , Y _s ) calculated by the read address generation circuit 14 to the video signal read from the frame memory 12, On the other hand, spatial image conversion (three-dimensional image conversion) desired by the operator can be performed. Therefore, the source video signal V stored in the frame memory 12. , And the object video signal V ₂ read from the frame memory 12 are two-dimensional data. That is, in涫算this two-dimensional readout completion de Les (X _s, Ys) is a three-dimensional space Above z-axis data is not used.

 Therefore, the parameters in the third row and third column for calculating the component in the Z-axis direction of equation (3) are not required when calculating the read address for the frame memory 12. .

Therefore, when the actual three-dimensional transformation matrix having para menu over data required for the operation of the two-dimensional readout completion dresses and T _33, the matrix T ₃₃ is

, (3), excluding the parameters in the third row and third column,

 Γ ΐ〗 Γ ΐ 2 Γ ΐ 31 Ζ

 丄 33 = Γ 21 22 Γ 231 Ζ (4)

 1 X Υ 1 2 Ρ Ζ + S

Can be represented by

 Here, the relationship between the position vector on the frame memory 12 and the position vector on the monitor screen 8 will be described.

In FIG. 5 lambda, the two-dimensional add-less on-off Re_厶Me mode Li _{1 2 (Χ Μ, Υ Μ} ) and the position vector preparative Le a [chi _Micromax Upsilon _Micromax], monitor scan click rie down 3 on § de-less (X _s, Y _s) and position the base click preparative Le and [Xs Y s]. The two-dimensional position base-vector on the full Remume mode Li 1 2 [chi _Micromax Upsilon _Micromax], is expressed in the same coordinate system, base click preparative Le [X _m y _m H. 〕 It can be expressed as. Further, when the Monitasuku rie down 5 position on the 5 base-vector [X _S Y _s] you expressed by homogeneous coordinate system can be represented as vector preparative Le [X ₆ y 6 1].

 The parameter “H.” in this homogeneous coordinate system is a parameter that represents the scale of the magnitude of the vector and the rate of reduction. Is used as pseudo depth information in the present embodiment. Rice H. Will be described later.

Nuclear and relative position vector on full Remume mode Li 1 2 [X _m y _m H _Q] and, by the action of the three-dimensional transformation matrix T ₃₃

, Full Remume mode Li 1 2 on the position base-vector [X _m y _m H. 〕 But It is converted to the monitor scan click position base-vector on rie down 3 [X _s ys 1]. Thus, full Remume mode Li 1 2 on the position base-vector [X _m y _m H. : And relationship between the monitor scan click position vector preparative le on rie down 3 [x _s ys 1], the following equation,

CX sys 1] = CX n, y _m H. ] · T _{3 3} · · · '(5).

The position vector on the frame memory 12 [X mym H. Parameter "H."及beauty homogeneous coordinate system used in], homogeneous coordinate system used Oite to position base click preparative Le on Monitasuku rie down 8 [X _s ys 1] The relationship with the parameter “1” is

, By a three-dimensional transformation matrix T _{3 3,} full Remume mode two-dimensional position downy click preparative Le on Li 1 2 [X _m y _m] of the secondary source position base click DOO on Monitasu click rie down 3 is converted to le [X sys], scaling ratio of the full Remume mode two-dimensional position downy click preparative Le on Li 1 2 [x _m y _m] "H." is the on Monitasuku rie down 3 follows represents be converted as scaling ratio of the coordinate system position base-vector of [X _s ys] becomes "1".

Thus, equation (5) is a relational expression for determining the point on the monitor scan click rie down 2 that corresponds to a point on the full Remume mode Li 1 2 Te matrix T _{3 3} Niyotsu. Here, in the image processing device using the special effect device, the source video signal is stored in the frame memory 12 before conversion, and is stored on the monitor screen ₃ obtained by the conversion matrix Τ33. The spatial video conversion is performed on the source video signal by determining the point of the frame memory 12 corresponding to the above point by the read address. In other words, when writing to frame memory 12

Instead of performing the image conversion, the image conversion is performed at the time of reading from the frame memory 12.

In such an image processing apparatus, the frame memory 12 Instead of calculating by the equation (5) to find the point on the monitor screen 3 corresponding to the point, find the point on the frame memory 12 corresponding to the point on the monitor screen 3 There is a need. Therefore, by transforming equation (5),

_{CX m ym H o = L xsy} s 1 ] ^{_{· Τ 3 3- 1 · '··}} , 6) by using a relational expression represented Te Niyotsu, off corresponds to a point on Monitasuku rie down 3 Remume A point on the memory 12 can be obtained. Therefore, according to the equation (6), the position base click preparative Le on Monitasuku rie down 3 [X _s ys 1] is specified, base position on the full Remume mode Li FM by the transformation matrix T ₃₃ one 'click DOO Le [x _m y _m H. ] Is calculated. Incidentally, the transformation matrix T ₃₃ - 'is the inverse matrix of the transformation matrix T _33.

Next, in order to obtain the two-dimensional position downy click preparative Le [chi _Micromax Upsilon Micromax] on full Remume mode Li FM, the transformation matrix T ₃₃ and the inverse matrix T ₃₃ - to ¹ as follows. That is, each element of the transformation matrix Τ ₃₃ is

 a, Si I 2 a 1 3

 a 2 & 22 α 23 (7) a 3 3 32 33

And the parameters of the inverse matrix T are:

X 33 ^{1 =}

 (8) a

 Where b

det (T ₃₃ )

As follows. Substituting equation (8) into equation (6),

[X my m H. ]

 = C b i i x s + b 2 1 y s + b a,

 b 1 2 X S 4-b 22 y S + b 32

b, 3 xs + b 23 ys + b ₃ 3] · · · · (9). Therefore, the following equation:

 X m = b i, X s + b 2 1 y S + b 3 I

 y m = b [2 X S + b 22 y S + b 32 (10)

 H o = b I 3 X S + b 23 y S + b 33

Is obtained.

Here, the position vector of the homogeneous coordinate system on the frame memory 12 [X my _m H. Is converted to a two-dimensional position vector [Χ _Μ Υ _Μ ] on the frame memory 12.

The parameter “Η.” Used to represent the two-dimensional position vector [Χ _Μ Υ „] in the homogeneous coordinate system is the parameter of the homogeneous coordinate system's position vector [X my _m }. Since it is a parameter that represents the scale factor, to convert the position vector of the homogeneous coordinate system to a two-dimensional position vector, the position vector of the homogeneous coordinate system must be converted. If the parameters “X _m ” and “y _m ” indicating the direction are normalized by the parameter “1-1.” Which indicates the scale of the position vector in the homogeneous coordinate system, good. Therefore, Monitasuku each parameter menu over data "X _M" in the two-dimensional position downy click preparative Le on rie down 3 and "Upsilon Micromax" is expressed by the following equation,

X _m

M ⁼

 H 0

 y m

Y _M = (1 1)

H o It can be expressed as. Monitasuku the homogeneous coordinate system vector preparative Le of the rie down 3 [x _s ys 1], in the same when converted into the two-dimensional position vector [X _s Ys], the position of the homogeneous coordinate system The parameters “X _s ” and “y _s ” indicating the vector direction are normalized by the parameter “1” indicating the scale factor of the position vector in the homogeneous coordinate system. Good. Therefore, the monitor scan click rie down 3 on a two-dimensional of each parameter "x _s" position downy click preparative Le and "Y _S" is expressed by the following equation,

 X s = X s

 Y s = y s (1 2)

It can be expressed as.

Therefore, the two-dimensional read address (X _M , Υ _Μ ) supplied to the frame memory 12 is given by the following equation from the equation (10).

X

 Μ ―

 H o

 b! 1 X s 4-b 2 1 y s + b 3 1

 b 1 3 X S + b 23 y S + b 33

 b, I xs + b 2 1 y s + b 3 1

 1 3) b I 3 X S b 2 3 y s + b

y

 Y

 H o

 b x + b 2 2 y s -I- b 32

 b 1 3 X S + b 23 Υ S + b 33

 b I 2 x s + b 22 ys 4-b 32

(1) can be obtained as a _{bi 3 xs + b 2 3 ys} + baa.

Next, each parameter bub ^ of T ₃₃ — 'is obtained. a 3 2 3.23 + fi 2 2 3.3

 b 11 (1 5)

 W,

 a 32 a 1 3 l l 12 a 23

 b I 2 ― (1 6)

 W I

 a 22 a. 13 + a l 2 a

 b, (1 7)

 w 1

 a 3 IS 23− 3-21 a 33

 b 2 I (1 8)

 w

 One a a + a a 33

b 22 ⁼ (1 9)

 W,

b 3 2 ― ( 2 2 )

b 3 2 ― (2 2)

 w, b 3 3 (2 3)

 W, becomes o where

 W 1 ~ ― 3.2 3 S 3 1 α. 1 3 + d 2 1 α 32 S 1 3 + 3 1 2 cl 3 I CI 23

 ― Tl l l S 3 2 3 23 ― ii I 2 cl 21 3.33 + S 1 I 3.22 S 33

· · · · (2 4) o Here, the values of au to a ₃₃ are, from the relationship of equation (7),

 a, i — 1 1, Ά 1 2 — T 1 2, a 1 3 = 1 3 Ρ ζ

3. 21 ⁼ 2 1, a 22— Γ 22, £ 1 2 3— Γ 2β Ρ ζ (26) a 31 = 1 χ, a 32 = 1 _y , a ₃ 3 = l _z Ρ ζ + s----(27), so by substituting this into the expressions (15) to (24),

 + Γ 2 1 1 y Γ 1 3 P Z

 Γ I 2 1 Γ 2 3 i Z

 1 Γ Γ 2 3 P Z

 ~ Γ, 2 Γ 2! (1 Z P Z + S)

+ r,, r ₂ 2 (lz P z + s)-(3 7

It can be expressed as.

Thus, by substituting the values of the expressions (28) to (37) into the expressions (18) and (14), the values are supplied to the frame memory 12. The read _address (X _M , Υ _Μ ) is

X Μ — C {-1 X 23 P z + r ₂₂ (lz P z + s)} X

 Η 0

+ {1 ₃ PZ + Γ, 2 (1 _ζ Ρ ζ + s)} Υ s

+ (One r ₁₃ Ρ ζ + r ₁₂ r ₂₃ Ρ ζ)]--

Υ C {1 χ Ρ ζ 一 2 I (1 ζ Ρ ζ + s)} X

 Η

+ {-1 _χ + r () ζ Ρ + s)} Υ

 + {Γ 2 1 Γ 1 3 Ρ-r 1 1 r Ρ ζ}] (given as 39. However, II.

 H o = (-2 2 X + r 2 1) y) A S

 + \. 1 2 1 X Γ 1 1 1 y

+ (− R _{l 2} r 2 1 + riir ₂₂ )…-(0)

It is. The read address (X _M , Υ „) supplied to the frame memory 12 is converted into a three-dimensional conversion matrix Τ determined by a desired spatial image conversion device of the operator. Each parameter (r,, ~ r

It can be expressed using the scan Bae Kuti blanking value _{P z - 33, 1 X,} 1 y, 1 z and s), and, Pas a parameter one data set in advance.

Therefore, when a screen address (Xs, Ys) to be addressed to the equations (6) to (40) so as to correspond to the raster scan order of the monitor screen 3 is supplied, the supply is performed. The read addresses (X _M , Y _M ) on the frame memory 12 corresponding to the screen address thus obtained can be sequentially calculated.

 As described in the HU, in this embodiment, the parameter 11 obtained by the expression (40) is used. Is used as depth information.

Because the Ζ ^ standard value of the actual 3D space when 3D conversion is performed H. Is proportional to

 Thus, the following effects are obtained by using the value of H obtained by the equation (40) instead of using the actual Z coordinate value in the three-dimensional space as the depth information. That is, when calculating the depth information, it is not necessary to calculate the actual Z coordinate value, so that the calculation for one dimension can be omitted. Thus, it is not necessary to use a high-speed processor for three-dimensional conversion, and depth information can be calculated even with a low-speed processor. In addition, this H. Since H is a value required when calculating the two-dimensional read address to be supplied to the frame memories 12 and 13, this H is used. There is no need to perform any special operations to find. Therefore, even faster calculation can be performed.

In this way, it obtained on the parameters H meaning of α Figure 6 space sump on the scan click rie down surface 5 5 Α Remind as in (Β) Li Ngua de-less (X _s, y s) of Assuming that the interval, that is, the interval between pixels on the screen surface 55 A is “1”, as shown in FIG. 6 (A), the read space of the frame memories 12 and 13 is It means that the interval between the sample addresses (X, y) is the value "H." In FIGS. 6 (A) and 6 (B), 〇 indicates data stored in the frame memory, and X indicates a read address point.

So, depth information Η. Parameter H as When the value is greater (1 3) and (1 4) normalized full Remume mode Li 1 2, 1 3 read add-less in formula _{(x M ZH., Y M} ZH.) Of The value becomes smaller. As described above, when the interval between the read addresses of the frame memories 12 and 13 is reduced, the screen surface 55

By increasing the number of pixels developed on A, the image projected on 55 A on this screen surface is enlarged.

On the other hand, no. When the frame size H decreases, the frame memory 1 2, 1 3 of the normalized read add-less _{(x M / HO, y M} ZH.) Means that the Naru rather large, thereby the spatial reading off Remume mode Li 1 2, 1 3 The interval between addresses increases. However, when the space between the spatial readout sampling addresses becomes large, the data read out from the frame memories 12 and 13 is displayed when displayed on the screen surface 55A. The resulting video is reduced.

 For example, as shown in Fig. 7 (B), in the address space of frame memories 12 and 13, parameters are set according to the perspective so as to intersect with the specified memory area ER1 without using the perspective. Data H. When the image data is read out by the read address of the area ER2 where the image is read out, the image projected on the screen surface 55A is projected in perspective as shown in Fig. 7 (A). The image data read out from the area ER1 of the frame memories 12 and 13 without using the image data is a spatial readout sampler determined by the same parameter H over the entire area ER1. By being read out by the dress, the image is displayed in the area ERIX on the screen surface 55A as a non-perspective image.

 On the other hand, the pixel data of the frame memories 12 and 13 are read from the area ER2 where the intervals of the spatial readout sampling addresses are different by the perspective method. Parameter H on lean surface 55A. In the read sampling address where is large, the image is reduced and the parameter H is set. In a region where the distance between the read sampling addresses becomes large and the interval between the read sampling addresses becomes large, a reduced image is displayed on the screen surface 55A.

In this way, the parameter H. Changes in the size of the image as depth information, so that images using perspective can be screened. Can be projected on the surface 55A.

 (4) Explanation of shadow coordinate system

Next, the shadow coordinate system will be described with reference to FIG. The shadow coordinate system, like the word Lumpur de coordinate system, X _s, is defined by three-dimensional orthogonal coordinate system consisting of YS and Z s axis is a coordinate system. As shown in FIG. 8, when the shadow given to the object video signal V ob j is a shadow video signal V shadow, the plane on which the shadow video signal V shadow is projected is defined as the XY plane of the shadow coordinate system. This is called the shadow plane. The direction in which the light source for providing the shadow of the object video signal Vobj exists is defined as the negative direction of the Z axis of the shadow coordinate system. As described below, the special effects device of the present invention generates a shadow using a point light source mode using a point light source and a shadow light source using a parallel light source. It has a parallel light source mode and can be freely set by the operator. Further, X in the whirlpool de coordinate system, against the Y and Z-axis, X s, the angle of each of the Y s and Z _s axis of shadow coordinate system can be O Operator sets an arbitrary angular .

 (5) Description of conversion processing for generating a shadow video signal in point light source mode

First, while參照Figure 9 A and FIG. 9 B, in case of the point light source mode, the conversion processing for obtaining the Shah Doubideo signal V ₄ converts the source Subideo signal will be described. 9A and FIG. 9B are the same as FIGS. 4A and 4B, and FIG. 9A is a view from the viewpoint PZ set on the Z axis of the world coordinate system to the XY of the world coordinate system. FIG. 9B is a diagram when the plane is viewed, and FIG. 9B is a diagram when the YZ plane in the world coordinate system is viewed from a position in the positive direction of the X axis in the world coordinate system.

First, as described with reference to FIGS. 4A and 4B, the three-dimensional transformation matrix T is used. 3D converted to 3D spatial position by The object video signal V, is perspective-transformed to the X s Y s plane of shadow coordinates by a perspective transformation matrix PSP _{ο τ} by a point light source. This is a video signal seen through the X _s Y s plane of the shadow coordinate system when the three-dimensional object video signal V, when viewed from the point light source 90, is viewed from the point light source 90. in _ς Figure 9 beta, which means to seek the video signal which is transparent to X s YS plane of shadow coordinate system, to be expressed as a three-dimensional Shah Doubideo signal V _3. A detailed description of the perspective transformation matrix PSPOT using the point light source will be described later.

Next, the three-dimensional shadow video signal V ₃ is converted to the perspective transformation matrix 説明 described above. Is used to perform perspective transformation so as to be seen on the XY plane of world coordinates. This is from a virtual viewpoint PZ on the Z-axis, when viewed three-dimensional sheet catcher Doubideo signal V _3, means to determine the video signal to be transparent to the XY plane of the whirlpool de coordinate system. In FIG. 9B, the video signal seen through the XY plane of the world coordinate system is represented as a two-dimensional shadow video signal V 4.

The above processing shown in Fig. 9B is summarized. In the case of a point source mode, conversion processing for the two-dimensional source over scan video signal V 0 obtains a two-dimensional Shah de video signal V ₄ are Obujuku preparative video signal V. To the 3D transformation matrix T. A three-dimensional conversion step for obtaining a three-dimensional object video signal V, by performing three-dimensional conversion with the three-dimensional object video signal V, and transforming the three-dimensional object video signal V, with a perspective transformation matrix P _{SP 0 T} using a point light source. a perspective transformation step of obtaining X s Y s perspectively in plan 3D Sha Doubideo signal V ₃ of the dough coordinate system, the three-dimensional Shah Doubideo signal V _3, perspective transformation matrix [rho. To obtain a two-dimensional shadow video signal V ₄ by seeing through the XY plane of the world coordinate system. Therefore, the two-dimensional source video signal V. From a two-dimensional shadow video signal The transformation matrix Tshadow 'for determining V 4 is given by

Γ shadow '= T ₀ -PSP OT * P o 4 1)

It can be expressed as.

 Here, the relationship between the position vector on the frame memory 22 and the position vector on the monitor screen 3 will be described. ,

The same as the relationship between the position vector on the frame memory 12 and the position vector on the monitor screen 3 is shown in FIG. 9C. the add-less _{(X M ', Y M'} ) and and position vector preparative Le _[Χ Μ 'Υ Μ'], the add-less on Monitasuku rie down 3 (X _s, Y s) and position Let the vector be [X _S Y _s ]. To express this full Remume mode two-dimensional position downy click preparative Le on Li 2 2 [XM 'Y _M'] in the same coordinate system, can be expressed as _{[x m 'y m' H s} ' ]. Also, when expressed position vector of on Monitasuku rie down 5 5 [X _S Y s] in the same coordinate system, can be expressed as pair-vector [x _s ys 1].

Here, the parameter [H _s ] of the homogeneous coordinate system is a parameter of the homogeneous coordinate system when the position vector on the frame memory 12 is represented by the homogeneous coordinate system. Similarly to the above, the parameter is a parameter representing the scaling factor of the size of the vector, and is used as pseudo depth information in this embodiment. It will be described later with respect to this "H _S".

The transformation matrix T shadow 'for obtaining the two-dimensional shadow video signal V 4 from the two-dimensional source video signal V 0 is a 4-by-4 transformation matrix, which will be described later. data of the third row and third column) is not substantially used and the frames on the actual three-dimensional transformation matrix having a two-dimensional readout add-less arithmetic required para main Ichita and T ₃₃ shadow ' The three-dimensional transformation matrix T ₃₃ shadow 'is applied to the position vector [χ · η' y _m 'H s] on the memory 22 to obtain the position on the frame memory 22. Vector [X _m 'y _m ' H s] is converted to a position vector [X sys 1] on monitor screen 3. If the relational expression is described below,

 C X s y s 1] = [X m 'y m' Hs] · Τ shadow '

 ····· (5),

The parameter “H _S ” of the homogeneous coordinate system used in the position vector [X _m 'y _m ' H s] on the frame memory 22 and the monitor screen position on the 5 relationship between vector [X sys 1] parameter used Te smell "1", the three-dimensional transformation matrix T ₃₃ shadow, frame Note Li 2 2, 2 2-dimensional on 3 position vector [X _m 'y _m'] is converted into two-dimensional position downy click preparative Le on Monitasuku rie down 2 2 [X _s y _s], the two-dimensional on frame Note Li 2 2 The scale factor “H _S ” of the position vector [x _m 'y _m '] is the scale factor of the position vector [X _s y _s ] of the homogeneous coordinate system on the monitor screen 22. Means that the rate is converted to be "1" o

Thus, (5) 'equation a point on the monitor scan click rie down 2 that corresponds to a point on the frame Note Li 2 2 matrix T ₃₃ shadow' is閲係formula for determining Te Niyotsu. The monitor screen obtained by the transformation matrix T ₃₃ stiadow 'is also used as in the case of reading out the data on the frame memory 12.

(3) Frame memory corresponding to the point on (2) By specifying the point on (2) by the read address, it is possible to perform aerial image conversion on the matte-processed source video signal. I'm sorry. That is, the image conversion is not performed when writing to the frame memory 22, but is performed when reading from the frame memory 22.

To find a point on frame memory 22 that corresponds to a point on monitor screen 3, use the following equation: [ _m 'y _m ' H _s ] = [xsys 1] ♦ (T ₃₃ shado ') ^-1

····· (6) '. Thus in accordance with this equation, monitor the scan click position downy click preparative Le on rie down 3 [X _s y _s 1] is specified, the transformation matrix (T ₃₃ shadow ') -' Therefore off Remume mode Li 2 2 Top position vector (x _m

'y _m ' Hs] is calculated. Incidentally transformation matrix _{(T 33 shad ow ') -} 1 is the transformation matrix T ₃₃ shadow' is the inverse matrix of.

Next, the perspective transformation matrix PSPOT based on a point light source when a point light source is used will be described with reference to FIGS. Figure 1 0 is the X _s axis direction of the shadow coordinate system Y _s Z s plane Figure der connexion when viewed, the point light source 6 0 3 dimensional Obujeku preparative video V, and the three-dimensional Shah Doubideo V ₃ Indicates the positional relationship. Incidentally, perspective transformation matrix PSPOT that by this point source, when the light source mode point using a point light source, a conversion matrix for obtaining the three-dimensional Obujiweku preparative Bido V, from the three-dimensional Shah Doubideo V _3.

 FIG. 11A shows a flow of the conversion process for the point light source 60, and FIG. 11B shows a flow of the conversion process for the object video signal.

 First, with reference to FIG.

In step SP1, the position of the point light source 60 represented by the world coordinate system is transformed into the shadow coordinate system by the transformation matrix F. The reason for this is that the perspective transformation matrix P _s will be described later in step SP5. 'Is not a perspective transformation matrix in world coordinates, but a perspective transformation matrix in shadow coordinates. Therefore, when the three-dimensional object video signal V, is perspectiveed on the X s Y s plane of the shadow coordinate system by the perspective transformation matrix P _so ′, the point light source 60 represented by the world coordinate system is used. The position needs to be converted to a shadow coordinate system. Here, this transformation matrix F- 'will be specifically described. First, the rotation angle of the X _s axis of the shadow coordinate system with respect to the X axis of the field coordinate system is X, the rotation angle of the Y _s axis of the shadow coordinate system with respect to the Y axis of the world coordinate system is ^ _Y , and Ζ of the world coordinate system is Let the rotation angle of the Z _s axis of the shadow coordinate system with respect to the axis be ^ 、, and let the origin of the shadow coordinate system be (X so, y so, Z so). Here, for the transformation matrix F— 'from the world coordinate system to the shadow coordinate system, the transformation matrix F from the shadow coordinate system to the world coordinate system, which has an inverse matrix relationship, is the product of the rotation matrix and the movement matrix. First, a transformation matrix F from the shadow coordinate system to the world coordinate system is obtained. The transformation matrix F from the shadow coordinate system to the world coordinate system is

F = R x (d> χ) ♦ R ν (/? V) · R _z (^ _z ) ♦ L (x so, yso »Z so]

(42) where the matrix R _x (0, the matrix RY (0v) and the matrix R _ζ ( _ζ ) are matrices for rotation,

 1 0 0 0

 0 cos θ χ s i η χ 0

R x (0 _x ) = (4 3)

 0 -s i π χ cos θ χ 0

 0 0 0 1 cos 0 γ 0 -s i η ^ γ 0

 0 1 0 0

 R Υ (θ ν) (4) sin 0 0 cos 0 γ 0

 0 0 0 1 cos θ s i η ^ 0 0

 s i η ^ cos 0 0 0

R ζ (θ _ζ ) (4 5)

 0 0 1 0

0 0 0 1 It can be expressed as. The matrix L (X y Z so) is a matrix for translation, and

0 0 0

 1 0 0

 L (Xy2so) = (4 6)

 0 1 0

 y z 1.

Therefore, since the transformation matrix F from the shadow coordinate system to the X oo 1-world coordinate system and the transformation matrix F— 'from the world coordinate system to the shadow coordinate system are in inverse relation, The transformation matrix F— ¹ is

F-'= L (X so, yso, zso) R _x- ' (^ x)

• R Y- ¹ () ♦ R: '(z)

-L (-x y so,-z s o)-R x (-0 x)

• R v (-Y) · R _z (-Θ z)

 ... (4 7)

It can be expressed as. Here, the matrix R _x (— _x ), the matrix R γ (-0 ν), and the matrix R _z (-θ _z ) are rotation matrices,

COS (-^ γ) 0 -sin (-(9 _Y ) 0

 0 1 0 0

R _v (-) (4 9) sin (-Y) 0 cos (-(9 _Y ) 0

0 0 ϋ 1

It can be expressed as. The row system UL (-Xso, one YZso) is a matrix for translation,

1 0 0 0

 0 1 0 0

 L (-X y o) (5 1)

 0 0 1 0

 -z. 1

In step SP2, as shown in FIG. 10, the position of the point light source 90 is moved to the position of the virtual point light source 91 on the _Zs axis by the translation matrix T _KS — '. The reason for this is that, in order to obtain the 3D shadow video V 3 for the 3D object video V, when viewing the 3D object video V, from the position of the point light source 9 o 0, the 3D object video V, The three-dimensional shadow video V ₃ can be obtained by making the video V, see through the X _s Y _s plane of the shadow coordinate system. However, in order to perform this perspective transformation process, the point light source to be the viewpoint must be located on the _Zs axis. Therefore, the translation matrix T _XSYS —! Moves the position of the point light source 90 parallel to the position of the virtual point light source 91 on the Z s axis.

と表すことができる。 Here, assuming that the coordinates of the point light source 90 set in advance by the operator are (XL, _{yL, zL} ), this translation matrix T _XSYS — ¹ becomes T

It can be expressed as.

 By step SP1 and step SP2, the conversion processing for the point light source ends.

It will now be described perspective transformation matrix PSPOT for producing 1 1 with reference to B in three-dimensional old Bujiweku Tobideo signal V, three-dimensional Shah Doubideo signal V ₃ from.

In step SP3, as in step SP1, the three-dimensional object video signal V, represented in the world coordinate system, is transformed into the shadow coordinate system by the transformation matrix F1 '. The reason for this is that the perspective transformation matrix P _s used in step SP5 described later. 'Is not a perspective transformation matrix in the world coordinate system, but a perspective transformation matrix in the shadow coordinate system. Therefore, the perspective transformation matrix P _s . , The three-dimensional object video V, is seen through the X s Ys plane of the shadow coordinate system, and the three-dimensional object video V, represented by the world coordinate system, is transformed into the shadow coordinate system. Need to be converted to In step SP4, as in step SP2, as shown in FIG. 10, the three-dimensional object video signal V, is converted to the X _s Ys plane of the shadow coordinate system by the translation matrix T _XSYS — ¹ . Move parallel to. In FIG. 10, the translated video signal is represented as a virtual three-dimensional object video signal V, '. The reason why the parallel movement, in step SP 2, the translation matrix T _XSYS - Yotsute in ', the position of the point light source 9 0, and translation so as to be positioned on the virtual light source 9 1 on Z _s axis So the three-dimensional object for this point light source 90 The translation matrix T _XSYS — ¹ so that the relative positional relationship between the video signal V, and the virtual object video signal V, 'relative to the virtual point light source 91 does not change. This is because the three-dimensional object video signal V, also needs to be translated. Next, in step SP5, the perspective transformation matrix P _{s is obtained} . 'By the, connexion, virtual 3-dimensional Obujiweku preparative video signal V,' and is fluoroscopy X s Y _s plane of shadow coordinate system. In FIG. 10, the video signal seen through the X s Y s plane of the shadow coordinate system is represented as a virtual three-dimensional shadow video signal V ₃ ′. This virtual three-dimensional shadow video signal V 3 ′ is, when the virtual three-dimensional object video signal V, ′ is viewed from the virtual point light source 91 with the virtual point light source 91 as a viewpoint, _{s The} video signal seen through the Ys plane.

Specifically, this perspective transformation matrix P _s . 'Is the perspective transformation matrix P shown in equation (2). From the relationship between and the viewpoint PZ, it can be considered that the viewpoint at the time of the fluoroscopy is at the position of the virtual point light source 91, that is, at the position of (0, 0, Z! _). The perspective transformation matrix P _{s is} obtained by replacing “P _z ” with “— 1 Z z _L ”. 'Is obtained. Thus, the perspective transformation matrix P _s . '

It can be expressed as.

Next, at step _SP6 , the virtual three-dimensional shadow video signal V ₃ ′ is moved in parallel to the X _s Y s plane of the shadow coordinate system using the translation matrix T _xsvs . In Figure 1 0, it will be expressed as 3-dimensional Shah Doubideo signal V ₃ video signals that have been translated. Figure 10. As can be seen, the three-dimensional shadow video signal V ₃ has a point light source 90 as a viewpoint, and when the three-dimensional object video signal V, is viewed from the position of the point light source 90, the three-dimensional shadow video signal V 3 a video signal which is transparent to X _s Ys plane. The reason for this movement is that in step SP 4, the three-dimensional object and the video signal V, were translated by the translation matrix T _xsys- ', so that the translation matrix T _XSYS ¹ This is to undo the parallel movement processing by.

Here, the translation matrix T _XSYS — 'and the translation matrix T _XSYS have an inverse relationship, so the translation matrix T _XSYS is

と表すことができる。

It can be expressed as.

Next, in step SP 7, converted by the transformation matrix F represented by Equation (4 2), a 3-dimensional shadow video signal V ₃ represented by shadow coordinate system, the whirl de coordinate system. As a result, a three-dimensional shadow video signal V 3 represented by a coordinate value in the world coordinate system can be obtained.

The processing shown in the flow of Fig. 11B is summarized. The processing for obtaining the three-dimensional shadow video signal in the world coordinate system from the three-dimensional object video signal V in the world coordinate system is performed by converting the three-dimensional object video signal V, in the world coordinate system from the shadow coordinate system. Step (SP3), and in the shadow coordinate system, project the 3D object video signal onto the Xs Ys plane of the shadow coordinate system to project the 3D shadow video signal in the shadow coordinate system. (SP 4, 3-5 and 3-6) and a 3D shadow in the shadow coordinate system 変 換 Converting the video signal from the shadow coordinate system to the world coordinate system (SP7).

Therefore, a perspective transformation matrix P _S by a point light source for obtaining a three-dimensional shadow video signal in the first coordinate system from the three-dimensional object video signal V in the world coordinate system. _PT is a transformation matrix F- ', a row shift matrix T _XSYS - ^1, and a perspective transformation matrix P _s . ', The translation matrix T _XSYS- ', and the transformation matrix F.

 丄] SPO T ^ F * TXSYS * Isο * ΓSYS * F ··· JJ).

Therefore, when this transformation matrix P _S0PT is substituted into equation ( ₄₁ ), the transformation matrix Tshadow 'for obtaining the two-dimensional shadow video signal V ₄ from the two-dimensional source video signal in the point light source mode is obtained. Is

 Γ shadow '= Τ. · P · P o

= T 0 * F * TXYS ~ '* P so' ♦ TXY s' ♦ F ¹ * V

 (56).

Here, on the basis of this transformation matrix Tshadow ', a method of calculating the readout add-less for reading Shah Doubideo signal V ₄ from the full Remume mode Li 2 2, based on the matrix Tobj represented by the formula (3) , is exactly the same as a calculation method for obtaining read add-less (X _M, Υ _Μ) for reading the Obujiweku preparative video signal V ₂ from the full Remume mode Li 1 2. In other words, the operation method is exactly the same as the operation shown in Expressions (3) to (14).

More specifically, this transformation matrix T shadow 'is a matrix of 4 rows and 4 columns, but a matrix excluding the components in the Z-axis direction (3rd row and 3rd column) as in equation (4). the 'and, the matrix T ₃₃ shadow' T ₃₃ shadow of the inverse matrix of (T ₃₃ shadow ') - ^one of the para main Ichita the following equation: bb _{1 2} 'b 1 ₃ '

(T ₃ 3shadow ') *' = b ₂₁ 'b 22' b ₂₃ '(5 7)

b 3 b 32 'b ₃₃ ' The read address of the shadow signal generator 20 and the read address supplied from the address generator 24 are (XM ′, YM ′). Referring to the calculation method from Expression (3) to Expression (14), this read _address (Χ _Μ ′, Υ _Μ ′) is expressed by the following expression.

It can be expressed as. Where lis is

H s = b 1 ₃ 'X + b 23 Ys 4 b 33'

Thus, full Remume mode Li 2 2 read A de Les supplied (XM ', Y _M') the three-dimensional transformation matrix Τ is Yotsute determined desired spatial image conversion processing of the operator. Each parameter of the _{(r,, ~ r 33,} 1 1 γ, y ζ and s), and, parameters are set in advance - Perth Bae Kuti blanking value is data [rho _zeta, the position of the point light source (x _L, y

Ζ), the rotation angle of each axis of the shadow coordinate system ( _χ , 0 V, 0), and the position of the origin of the shadow coordinate system (x _s , y _s , z _s ).

Thus (6) to (4 0) expression for a monitor scan click rie down 3 for A Doretsu sheet ring to correspond to Rasutasukiya down order of the scan click rie N'a de Les (X _s, Y _s ), The read addresses (Χ _Μ ', Υ _Μ ') on the frame memory 22 corresponding to the supplied screen address can be calculated in order. You. As a result, a two-dimensional shadow video signal V ヽ corresponding to the two-dimensional object video signal V ₂ can be generated. (6) Description of conversion processing for generating a shadow video signal in the parallel light source mode.

First, referring to FIG. 1 2 A and FIG. 1 2 B, when the collimated light source mode one de using collimated light pi, 3D Obujiweku Bok video signal V, for obtaining the three-dimensional Shah Doubideo signal V ₅ from The conversion process will be described. 12A and 12B are the same as FIGS. 9A and 9B, and FIG. 12A is a view from the viewpoint PZ set on the Z-axis of the world coordinate system. FIG. 12B is a view when the plane is viewed, and FIG. 12B is a view when the YZ plane in the world coordinate system is viewed from the position in the positive direction of the X axis in the world coordinate system.

First, a three-dimensional transformation matrix T. The 3D object video signal V, which is converted to a 3D spatial position, is projected onto the X s Y s plane of the shadow coordinates by the perspective transformation matrix 平行_{Ρ Λ に}よ_{る with} a parallel light source Is done. In Figure 1 2 B, the X s Y _s perspective video signal to the plane of the shadow coordinate system, would represent by a three-dimensional Shah Doubideo signal V _5. The directional light based on the perspective transformation matrix P _PARA is the three-dimensional object video signal V, and the perspective transformation, a transformation matrix for obtaining the three-dimensional Shah Doubideo signal V _5.

Next, three-dimensional Shah Doubideo signal V ₅ is a perspective conversion matrix described earlier [rho. Is projected onto the XY plane of the world coordinate system. This means that when viewing the three-dimensional shadow video signal V 5 from the virtual viewpoint PZ on the Z axis, a video signal seen through the XY plane of the world coordinate system is obtained. In Figure 1 2 B is a video signal which is transparent to the XY plane of the whirlpool de coordinate system, to be expressed as a two-dimensional Shah Doubi de old signal V _6.

The above processing shown in Fig. 13B is summarized. Parallel light source In mode, two-dimensional source video. Signal V. Conversion processing for obtaining the two-dimensional sheet catcher Doubideo signal V ₆ from the source Subideo signal V. To the 3D transformation matrix T. 3D conversion step for obtaining a 3D object video signal V, by performing a 3D transformation by the 3D object video signal V, and converting the 3D object video signal V, into a shadow coordinate system by a perspective transformation matrix P _PARA using a parallel light source A perspective transformation step for obtaining a three-dimensional shadow video signal V ₅ by looking through the X s Ys plane of the three-dimensional shadow video signal V ₅ ; The result perspectively the XY plane of the world coordinate system, comprised of a step for obtaining a two-dimensional Shah de video signal V _c. Therefore

, Two-dimensional source video signal V. Transformation matrix Tshadow〃 for obtaining the two-dimensional Shah Doubideo signal V ₆ from the following equation,

T shadow "= T. · P _PA RA · P o ··· (60)

It can be expressed as.

The relationship between the position vector on the frame memory 22 in the parallel light source mode and the position vector on the motor task line 3 is similar to the relationship in the light source mode described above. shadow signal generator 2 0 read § de-less generation circuit 2 4 respectively read § de-less and position downy click preparative Le supplied from the (X _M 〃 Upsilon _Micromax 〃) and [monitored with Χ _{Μ 'Υ} "' scan click rie down 3 § dress and position base-vector in on the respective (X _s

Y s) and when the [X _s Y s], respectively when representable off Remume mode Li 2 2 and Monitasuku rie down 3 on the position base-vector at each homogeneous coordinate system [x _m "y _m "it is revealed tables and H _s) and (x _s ys 1].

For from the binary source Subideo signal V 0 to obtain a two-dimensional Shah Doubideo signal V 4, as will be described later the actual transform matrix having the parameters necessary for the calculation of two-dimensional § address reading T ₃₃ shadow When expressed as 〃, the homogeneous coordinate system position vector on this frame memory 22 闋係Le [xm "y _m 〃 H s] and the monitor scan click homogeneous coordinate system position vector preparative le on rie down 3 [X _s y _s 1] is expressed as follows.

[X sy _s 1] = [X _m "y H s] '· T ₃₃ shadow'

 (5) 'Therefore, to find the point on the frame memory 22 corresponding to the point on the monitor screen 3,

[X _m "y H _s ] = [xsys 1] · (T ₃₃ shadow ')-'

 ····· (6) '.

Next, the perspective transformation matrix P _PARA by the parallel light source when the parallel light source is used will be described with reference to FIGS.

 FIG. 13A shows the flow of the conversion process for the parallel light source 70, and FIG. 13B shows the flow of the conversion process for the three-dimensional object video signal V ,.

 First, with reference to FIG. 13A, a description will be given of a conversion process for the parallel light source 70.

In step SP11, the coordinates of the parallel light source 70 defined by the spherical coordinates in the world coordinate system are converted to the orthogonal coordinates in the world coordinate system. The position of a parallel light source is generally expressed in a spherical coordinate system rather than in a rectangular coordinate system. The spherical coordinate system is a coordinate system in which the position of a parallel light source is represented by “radius (r)”, “latitude (a)”, and “longitude (β)”. Figure 14 1 shows the relationship between rectangular coordinates and spherical coordinates in the world coordinate system. As shown in Fig. 12A, the relationship between the rectangular coordinate system and the spherical coordinate system is that the reference for latitude () is the ii direction of the Y axis and the reference for longitude is the XY plane. In other words, the equatorial plane of the spherical coordinate coincides with the XY plane of the rectangular coordinate, and the direction of latitude 0 (rad) and longitude 0 (rad) coincides with the minus direction of the Y axis. Therefore, the position of the collimated light source 70 defined by spherical coordinates is (r, a, β), and the position of the collimated light source converted to rectangular coordinates is (χ ι_, y L, z L), the light source position (XL, y L, z _L ) is

It can be expressed as. Therefore, in order to convert the light source position defined by spherical coordinates into rectangular coordinates, the light source position (r, α, β) of the spherical coordinate system set by the operator is substituted into this equation (61). Thus, the light source position (XL, yL, Z! _) Converted into the rectangular coordinates can be obtained.

In step SP12, the coordinates of the parallel light source are transformed from the first coordinate system to the shadow coordinate system by the transformation matrix F- '. For the transformation matrix F one ^1, since as described in step SP 1, description is omitted. Shadow coordinate system the position of the converted light source (x i_ ', y L' , z L ') and when, homogeneous coordinates source base click preparative Le of in the world coordinate system expressed in [XL y _L z _L 1] and the vector of the light source in the shadow coordinate system represented by the homogeneous coordinate system [XL _L y L L z _{L L}

1] and the transformation matrix F— ¹ are as follows:

 C L 'y L z L 1]

= [XL y L z _L n ♦ F-'· · · · · (6 2).

In step SP13, the position of the light source in rectangular coordinates in the shadow ゥ coordinate system (X, y, yL ', z') obtained in step SP12 is expressed in spherical coordinates in the shadow coordinate system. Convert to represent. Ei 14B is a diagram showing the relationship between the rectangular coordinates and the spherical coordinates in the shadow coordinate system. As shown in Fig. 14B, the relationship between the rectangular coordinate system and the spherical coordinate system in the shadow coordinate system is such that the reference of the latitude (ar _s ) is the negative direction of the Y _s axis and the reference of the longitude (β _s ). And the X _s Y _s plane are doing. That, X s of the equatorial plane is orthogonal coordinates of the spherical coordinates - matches Y s plane, latitude 0 (rad), the direction of柽度0 (rad) matches the minus direction of the Y _s axis.

 Therefore, the position of the light source (X ι_ ', yz L') expressed in the rectangular coordinate system in the shadow coordinate system and the position of the light source (r ', α', β ') expressed in the spherical coordinate system are calculated. The relationship is

r '= Sqrt {(X + (y) ² + (z

a '= tan " ¹ (— _L ', y L ')

β '= sin- ¹ (— z L', r)

····· (6 3) Therefore, the position of the light source obtained in step _{SP 1 2 (X y L '} , z') and by Rukoto be substituted into the equation (6 3), the position of the collimated light source 7 0, the shadow coordinate system It can be expressed in spherical coordinates.

Next, in step SP 14, as shown in FIG. 15, the position (r ′, a ′, β ′) of the parallel light source obtained in step SP 13 is calculated by using the rotation matrix R _z (Ichi '). ) Rotates one "" (rad) around the 豳_s axis of the shadow target system. In other words, the position of the parallel light source after the rotation process can be expressed as (Γ ', 0, β'). In Fig. 13, the light source rotated by the rotation matrix R _z (~ α ') is represented as a virtual collimated light source 71. By rotating the position of the collimated light source Ί0 in this manner, the light source shown in Fig. 13 is obtained. As shown, the luminous flux of the parallel light source Ί 1 is parallel to the Y _s Z s plane, and the reason why the high-speed of the parallel light source 7 1 is rotationally converted to be parallel to Y s Z s is described later. In step SP 17, the observable video signal V, is perspectiveed on the X _s Y _s plane of the shadow coordinate system by the perspective transformation matrix P so '. If the incident luminous flux of the collimated light source is parallel to the YsZs plane, then even if the 3D object video signal is seen through the XsYs plane, Since the coordinate values in the Xs-axis direction do not change, a perspective transformation matrix P _s for allowing the object video signal to be seen through the X _s Ys plane. 〃 can be represented very easily.

Specifically, since the rotation matrix R _z (-α ') is a rotation matrix around the Z _s axis, the rotation matrix R _z () is

 (64).

 By the steps SP 11, SP 12, SP 13, and SP 14, the conversion process for the parallel light source 70 is completed.

Next, referring to FIG. 13B, in a parallel light source mode using a parallel light source, a perspective transformation matrix for generating a three-dimensional shadow video signal V ₅ from the three-dimensional object video signal V, I will explain P _PARA .

In step SP15, similarly to step SP11, the three-dimensional object video signal V, represented in the world coordinate system is transformed into the shadow coordinate system by the transformation matrix F '. The reason for this is that the perspective transformation matrix P _s , which will be described later, is similar to the processing for the parallel light source described in step SP 11. This is because 〃 is not a perspective transformation matrix in the world coordinate system, but a perspective transformation matrix in the shadow coordinate system. If the 3D object video V, is to be seen through the X s Ys plane of the shadow coordinate system, each pixel position of the 3D object video V, represented by the world coordinate system, is shadowed. Need to convert to coordinate system. In step SP 16, the object video signal V, converted to the shadow coordinate system in step SP 15, is rotated by one “” (rad) around the Z axis by the rotation matrix R _s (-'). This is because, as shown in FIG. 15, in step SP 14, the position (r ′, α ′, β) of the parallel light source 70 is determined by the rotation matrix R _z (— “”).

This is because the object video signal V, has to be rotated to correspond to the rotation process of the parallel light source 70, since ') has been rotated about the Ζ axis — “' (rad). As shown in FIG. 15, the three-dimensional object video signal that is rotated (rad) around the Z _s axis by the rotation matrix R ₂ (-a ') is converted into the virtual three-dimensional object video signal V, ". Therefore, the relative position of the three-dimensional object video signal V and the parallel light source 70 with respect to the origin of the shadow coordinate system, and the virtual three-dimensional object video signal V with respect to the origin of the shadow target system , 〃 and the virtual parallel light source 71 are exactly the same.

Next, the stearyl-up SP 1 7, Te perspective transformation matrix P _s Niyotsu, virtual 3-dimensional Obujiweku preparative video signal V, and 〃 and fluoroscopy X _s Y _s plane Shah de coordinate system. The perspective transformation matrix will be described with reference to FIGS. 16A and 16B. First, as shown in FIG. 15, FIG. 16A and FIG. 16B, the virtual object video signal V, "

, The perspective transformation matrix P _s . The video signal seen through the X _s Ys plane by 〃 is represented as a virtual three-dimensional shadow video signal V ₅ ′. FIG. 14A is a drawing that three-dimensionally represents the positional relationship between the virtual three-dimensional object video signal V, ′ and the virtual three-dimensional shadow video signal V ₅ ″, and FIG. , the virtual 3-dimensional Obujiweku preparative video signal V when the positive direction of X _s axis viewed Y _s Z s plane, a ", a drawing representing the position閲係the virtual three-dimensional Shah Doubideo signal V ₅ 〃 is there. Here, the virtual object video signal V, 'J- Let a pixel point with (x., _Y. , ¾.) Be a perspective transformation matrix P _{s of} this pixel (., Yo, zQ). 〃 by Xs Ys perspective the pixel point on the virtual Shah Doubideo signal V ₅ to the plane (X _s, ys, zs) When, from the geometric relationships shown in FIG. 1 6 B,

 X s = X 0

 y s = yo + z 0 cot β '

It can be seen that the relationship zs-0 · · · · (6 5) holds. The virtual Obujeku preparative bi de O signal V, 〃 point above _{(x., Y Q, z} .) And virtual Shah Doubi de O signal V ₅ on the point (x _s, ys, zs) and the perspective transformation matrix P _s . Is the following equation

X syszs 1] = X ₀ yozo 1] · P so '··· ^ ^ 6). Therefore, from equations (65) and (66),

1 0 0 0

 0 1 0- 0

 P s o '

 0 cot β '1 0

 0 0 0 1 (67).

Next, in step SΡ18, virtual shadow video signal V ₅回 転 is rotated around Z _s axis by rotation matrix R _z { ¹ ). In FIG 5 depicts a rotational movement video signals as a three-dimensional sheet catcher Doubideo signal V _5. As can be seen from FIG. 15, the three-dimensional shadow video signal V ₅ is a video signal obtained by seeing the three-dimensional object video signal V, through the parallel light source 70 into the X _s Y s plane of the shadow coordinate system. It is. The reason for moving this way

, In step SP 16, the rotation matrix R _z (— “') gives the third order This is because, since the original object video signal V, has been rotationally moved, the rotational movement processing by the rotational matrix R _z {-a ') is restored. In other words, since the rotation matrix R _z ( ^-ar ) and the rotation matrix R _z (α ') have an inverse matrix relationship, the rotation matrix R _z (α') is

 (68).

Next, in step SP 1 9, converted by the transformation matrix F represented by Equation (4 2), a 3-dimensional shadow video signal V ₅ represented by shadow coordinate system, the whirl de coordinate system. As a result, a three-dimensional shadow video signal V5 represented by the coordinate values in the world coordinate system can be obtained.

1 3 when B summarize the process shown in flow of, the process for three-dimensional Obujiweku preparative video signal V Wa Lumpur de coordinate system, obtains a three-dimensional Shah Doubideo signal V ₅ of the world coordinate system, 3 Step (SP15) for transforming the three-dimensional object video signal V, from the world coordinate system to the shadow coordinate system, and converting the three-dimensional object video signal into the Xs of the shadow coordinate system in the shadow coordinate system. Steps (SP1ΰ, SΡ17 and SΡ18) for projecting onto the Ys plane to generate a three-dimensional shadow video signal in the shadow coordinate system, and three-dimensional shadow video in the shadow coordinate system A step (SP 19) for converting the video signal from the shadow coordinate system to the world coordinate system to obtain a three-dimensional shadow video signal V 5 in the world coordinate system.

Therefore, whirlpool de coordinate system of the three-dimensional Obujiweku Tobideo signal V, from the determined three-dimensional Shah Doubideo signal V ₅ whirlpools de coordinate system The perspective transformation matrix P _PARA is a transformation matrix F _ ′, a rotation matrix R _z (—), and a perspective transformation matrix P _s . 、, the rotation matrix R _z ) and the transformation matrix F— ^1.

P PARA = F-'♦ R a {-')-P so '· R _z () · F

 ····· (6 9)

Therefore, when this transformation matrix P _PARA is substituted into equation (60), the transformation matrix Tshadow〃 for obtaining the two-dimensional shadow video signal V ₄ from the two-dimensional source video signal in the parallel light source mode is obtained. Is

T shadow "= Τ. · P P ARA · P o

Two T. ♦ F-'· R — · P _s · R _z («') · F · P

 ····· (70)

Here, on the basis of this transformation matrix Tshadow〃, show two-dimensional Shah de Ubideo signal V ₆ calculation method for calculating the add-less read for reading from off Remume mode Li 2 2, the formula (3) based on the matrix T obj to be a two-dimensional Obujiweku Tobideo signal V ₂ the full rate厶Me motor-less read add-in for reading from Li _{1 2 (X M, Υ Μ} ) is exactly the same as潢算method for determining the . That is, from equation (3),

 The calculation method is exactly the same as the calculation shown in (14).

Specifically, since this matrix T shadow "is a 4-by-4 matrix, the matrix excluding the components in the Z-axis direction (the third row and the third column) is calculated as in Equation (4). and T ₃₃ Shadow〃, inverse matrix of the transformation matrix T ₃₃ Shadow〃 (T ₃₃ shadow ") each parameter one ^1, the following equation

 1 Γ b i 2 b I a

 (T 33snadow ") b 2 I b 22 'b

baib 32'b (set to 7 1. Also, read address generation of shadow signal generator 20—- Let the read address supplied from the circuit 24 be (Χ Μ ", YM '). Referring to the calculation methods from Expression (3) to Expression (14), this read address (X _M ，, Υ _Μ ") is the following equation: b X s + b 21 Ys + b 31

 X (72)

b X s + b 23 Ys + b 33 b X s + b ₂₂ 〃 Y s + b

 Y M "= (7 2 ')

b: 'Xs + b23Ys + b33. Where H _s ' is

H s = b X s + b ₂₃ Y s + b 33 "

Therefore, read A de Les supplied to off Remume mode Li _{2 2 (X M ", Υ} Μ 〃), and each of the three-dimensional transformation matrix T. is Yotsute determined desired spatial image conversion processing operator Parameters ( _{Γ ι 1} Γ Γ

33. 1 X, 1 V, 1 2 and S), and, Perth Bae Kuti blanking value [rho ₂ is a parameter one data set in advance, the position of the point light source (XL, y L, z t_ ), shadow coordinates rotation angle of each axis of the system _(χ, Θ V, the position of the origin of theta Shah de window coordinates _{(X s, y so, z} s;..) can be represented using.

Thus (6)对the formula - (4 0) type, scan click rie to § Doretsu sheet ring to correspond to Rasutasukiya down顺of Monitasuku lean 3 N'a de Les (X _s, Ys) supply then, the supplied scan click rie N'a de-less off corresponding to the Remume mode Li 2 add-less out read on the 2 (X _M 〃, Y ""), and Ru can be calculated in the order. This Yotsute can generate two-dimensional Shah Doubideo signal V ₆ corresponding to two-dimensional object video signal V _2.

 (7) Explanation on setting of shadow coordinate system

As described above, a shadow coordinate system for defining a shadow plane on which a shadow is projected for an object video signal is set. The rotation angles of the Xs, Ys, and Zs axes of the shadow coordinate system with respect to the X, Y, and Z axes of the world coordinate system

_{X, 0 γ, θ z,} , one -.) TOLEDO coordinate origin position of the origin of the shadow coordinate system with respect to the (X _S, y so, it is necessary to set the Z so). In the special effect apparatus of the present invention, as described above, so that the operator sets the X s axis, the origin of the rotation angle and shea catcher dough coordinate system Ys axis and Z _s axis of arbitrarily Shah de window coordinate system It has become. Equation (4 2)及beauty expression for (4 7), the rotation angle of each axis _{(<9 χ, θ V,} ^ z) and the origin of the coordinate _{(x s., Y so,} z s.) Assignment Is done. However, if the operator sets an appropriate position as the origin of the shadow coordinate system in this way, as shown in FIG. 15A, the three-dimensionally transformed object video signal V, and the three-dimensional shadow video signal V _{5 become} However, they may be spatially separated. This is because the 3D object video signal V, does not exist on the shadow plane, in other words, the 3D object video signal V, does not exist on the shadow plane. It means that the shadow plane cannot be set as it exists. Of course, there is no problem if the operator wants a shadow surface as shown in Figure 17A. However, in order for the light from the light source to have the effect of shadowing a certain object existing on the ground, it is necessary to set this ground as the shadow plane on which the shadow is projected. There is. That is, it is necessary to set the origin of the shadow coordinate system so that the three-dimensional object video signal V, spatially exists on the shadow plane. For this purpose, the special effect device of the present invention has an origin setting mode for automatically setting the origin of the shadow coordinates.

First, the operator sets the source video signal V. Specify a point above. 3D transformation of this specified point The object is converted into a three-dimensional space by the matrix T _Q , and the corresponding point on the object video signal Vobj converted into the three-dimensional space is set as the origin of the shadow coordinate system. As a result, the origin of the shadow coordinate system is set at a point on the three-dimensional object video signal Vobj, and as a result, the three-dimensional object video signal Vobj exists on the shadow plane. become.

Specifically, as shown in FIG. 17B, the source video signal V. Let the upper right point be a, the upper left point be b, the lower left point be c, and the lower right point be d, and the corresponding point on the 3D object video signal V ob j be a ' , b ', c', and d '. Also, the source video signal V. The four points a, b, and each of the coordinates of c and d, the (X a, ya, 0) , (x b, y b, 0), (xc, y c, 0)及Beauty (x _d, and y _d, 0), a point on the 3-dimensional Obujiweku preparative video signal V ob j ', b', c ' and a' respectively of _{coordinates, (x a ', ya'} , za ') (xb' _{, yb ', z b')} , and _{(x c ', yc',} z c ') and _{(x d', y d '} , z d').

Next, the source video signal V is output by the operator. The case where the above point d is specified is explained as an example. Source video signal V specified by the operator. The corresponding point on the observable video signal V obj obtained by transforming the above point d (X _d , y _d , 0) by the three-dimensional transformation matrix T o is represented by d ′ (X c ′, y _d ′, z _d '). This point, d (x _d , yd, 0) and d '(

 ') And the vectors of the homogeneous coordinate system, respectively.

C DYD 0 1], the vector preparative consisting Le _{[x d 'y "' z d} '1 ], and these vectors,闋係the three-dimensional transformation matrix T. The following equation, C x ό 'y ά' z A '1] [dyd 0 1] T

C xdy _d 0 1]

 -... (73).

 Therefore, from this equation (73),

X d '2 (τ 1 1 X d + r 2 1 y d + 1 x) / S

y _d '= (r 1 2 xo + r 22 yd + 1 _Y ) / s

_{z d '= (r 1 3} xo + r 2 ayd + l z) / s · · · * (7 4) relationship is established that.

Therefore, Equation (4 2) and the origin of Shah de coordinate system is a path lame one other mobile matrix L represented by the formula (4 7) (X _s., YSO, ZSO), a in Equation (7 4) connexion obtained _{(x d ', y d'} , z d ') by substituting, 3-dimensional objects Tobideo signal V, the origin of the shadow coordinate system to a point above is set. In other words, the shadow coordinate system is set so that the three-dimensional object video signal V, exists in the shadow plane in the three-dimensional space. Accordingly, the special effect device of the present invention can obtain a natural effect such that the shadow of an object existing on the ground is projected on the ground by the light from the sun.

 (8) Description of real-shadow generator 50

This Kodefu Remume mode Li 2 3 Sha Douki signal output from the (at Tenko猄mode is K _4, when directional light mode is kappa ₆₎ is entered, a real for this key signal shadow, ie re Arusha de one for outputting Riarushi catcher synchronizing first signal for generating and outputting a Mr shading and Boca (time point source mode one de the kappa ₄ ', when directional light mode one de the kappa _6') - The ゥ key generator 5 し な が ら will be described with reference to FIG.

 The real shadow generator 50 includes a gain control circuit 500, a horizontal LPF 501, a vertical LPF 502, and a real shadow control unit 503 as a whole. Also, the real shadow control section 503 includes a gain characteristic of a gain control circuit 500 described later, and horizontal and vertical

ROM 504, which stores table data indicating the filtering characteristics of 01 and 502.

The gain control circuit 500 is a circuit that controls a gain given to a shadow key signal (# ₄ , # 6) supplied from the frame memory 23. Horizontal and PF 5 0 1 GMBH Douki signal (Κ _4, Κ 6) from the gain control circuit 5 0 0 mouth with respect to the horizontal direction of the frequency component of - a path off I filter (LPF). Vertical LPF 5 0 2 is a horizontal LPF 5 0 Shah synchronizing first signal from 1 (Κ _4, Κ ₆₎ mouth one path off I filter in the vertical direction of the frequency component of (LPF). The real-shadow controller 503 receives a parameter と used as depth information in the present embodiment from the read address generation circuit 24 記憶 and stores it in R〇Μ 504. The gain control circuit 500 and the horizontal and vertical PFs 501 and 502 are controlled based on predetermined gain characteristics and filtering characteristics, respectively. Here, the predetermined gain characteristic is data that results in the intended gain characteristic, and the filtering characteristic is data that results in the intended filtering characteristic.

 In this embodiment, the final image to be obtained is such that the further away the shadow S from the object 0 is from the object に, the thinner and blurred the shadow S itself is, as shown in FIG. 19 (Λ). is there.

Note that this shadow does not change smoothly in the gradation area for the sake of drawing, but it actually changes very smoothly.

The gain characteristics (GAI を) indicating the shading of the shadow S are shown in Fig. 19 (B). As shown in Fig. 19 (B), the depth information As parameter H is large Kunar used, the gain is lowered against Shah Douki signal (K _4, K _G), as para menu over data H is rather small, the gain for the Shah Douki signal (K _4, KG) Can be raised. In other words, the brightness decreases as the shadow S moves away from the object 0. Here, as described above, the characteristic data shown in FIG. 19 (, 、) is shown in FIG. 18. The table data consisting of the lame signal and the gain data is shown in FIG. It is stored in R 0 Μ504 in the control unit 503.

Also, the blur amount (softness) of the shadow S, that is, the filtering characteristic (FILTER) of the one-pass filter is as shown in FIG. 19 (C). As shown in Fig. 19 (C), as the parameter Hs increases, the low-pass bands of the horizontal and vertical low-pass filters 501 and 502 become narrower, and the parameter Hs increases. H _s hands horizontally and vertically as the is rather small lowpass off I filter 5 0 1 5 0 of each low frequency 2 pass full I Rutagafu rats, i.e. wider. That is, the further away the shadow S is from the object 〇, the more the outline of the shadow S is blurred. The characteristic shown in FIG. 1 9 (C), para menu over data H _s passband comprising a full I filter coefficient data to variable Tepurude Ichita is shown in FIG. 1 8 as described above It is stored in R〇M504 of the real shadow control unit. One type of characteristic data shown in FIG. 19 (C) is used for the horizontal and vertical one-pass filters 501 and 502 shown in FIG. The use of a horizontal-port one-pass filter 501 means that high-frequency components cannot pass in the horizontal direction, so the less high-frequency components, the less distinctive the contours, etc. . The same applies to the vertical direction.

 (9) Description of combiner 30

Next, the explanation will be made with reference to Fig. 20 with reference to the combiner 30, but before that, the general synthesis theory of the combiner 30 will be explained. To

Generally the two video signals VA and V _B, respectively on the basis of the keys signals KA and K _B, will be described to be synthesized in accordance with priority as determined by the depth information H _0A. With respect to the first video O signals V _A, synthetic formula in the case of the second video signal V _B is above (closer to the disk re Ichin) is

VA _{(0 U} T)-KAVA + (1-KA) KBVB ··· (75 ₎ , and a composite output like V _{0 (0 UT)} is obtained.

Similarly, when the first video signal _VA is above the second video signal VB, the combining equation is

VB (OUT) = KBVB + - represented by _{(1 K B) KAVA ·· ··} (7 6), combined output such as V _{B (0 UT)} is obtained.

The first video signal V _A, in the case where the second video signal V _B is above (disk near rie down), priority signal to the composite video signal _{V A (0U τ)} Η 0Α Taking into account, the composite video signal corresponding to the first video signal V _A is

 V A (0 UT) X ― V A (OUT) H oA (7 /)

The following equation can be obtained by substituting equation (75) into equation (77).

ν _{Α (} ουτ) χ- {K AV A + (1-K _A ) K BV B} Η ΟΑ ···· (78)

The second video signal V _B, prioritization Li Ti signal when the first video signal VA is above can be represented by using the priority signal Ho A and (1 one H. _A). Therefore, when the first video signal V _A is above the second video signal V _B and the priority signal (111 H _0A ) is considered for V _{A (0 U} ), The output of V _B is

VB (0 UT) X ― VB (0 UT), 1 ― Η θΑ And by substituting equation (32) into this,

 V B (OUT) X = (K B V B + (1- K B) K A V A} (1 ~ H OA)

 ····· (80)

So the final composite video output V. _UT is

 V 0 U Τ Κ 0 U Τ = V A (0 U Τ) X + V B (0 UT) X (8 1)

If the keying process is performed with the key signal K OUT synthesized as shown in FIG. _UT is:

 V A (OUT) X + V B (0 U T) X

 V OUT = (8 2)

 A 0 τ τ

Becomes

Meanwhile, the first and second video signals V _A and V _B to E Li other than E Ria that image corresponding is displayed Ryo (i.e. We Ria neither is displayed video signal VA or V _B) is, using the key signal K _a and K _b of the first and second video signals VA and V _B (1- K _a) - so can be defined in the form of a product such as (1 K _N), the video signal V _Alpha Or the key signal OUT for the area where either VB or

K OUT = 1-(1-KA) (1 K _B ) (83)

Can be represented by

Therefore, by substituting equations ( ₇₈ ), (80) and ( ₈₃ ) into equation ( ₇₇₎ , the composite video output V _ουτ finally obtained is V 0 u '({K _A V _A + (1-K _A ) KBV _B } H OA

 + {K B V B + (1-K B) K A V A} (1-II ΟΛ)

 1

 ({H OAK A + (1-H OA) (1-K))} V A

+ {(1-H OA) K B + H OA (1 — K) K} V B)

 1

 X (84)

It looks like 1 1 (1 1 KA) (1-KB). Therefore, the key signal で expressed by equation (83) and equation (84). _υτ and the composite video output V _0UT are output from the _{compiler 30} as a mixed key signal K ′ _mi „and a mixed video signal V ′ _miK .

The combiner 22 of the special effect device of the present invention has two pieces of image information, that is, depth information II in the XYZ space behind the screen surface 3A shown in FIG. 20 (A). And two video signals of images intersect each other while extending obliquely so as to have a H _s, Sunawa Acquired Bujiweku preparative video signal V ₂ and Shah de video signal (point source mode at the time of V _4, directional light mode time for V ₆₎ de, respectively the key signal K ₂ (FIG. 2 0 (B)) and K ₄ (point source mode when the kappa _4, when directional light mode is K s) (Fig. 2 0 (C) while keys Lee in g by), priority signals of two images each part Ho _a and (1 one H. _a) (Fig. 2 0 (D) Figure based on and FIG. 2 0 (E)) 2 1 Are synthesized using the configuration of

Here, the prioritization Li Ti signal H OA, is information for indicating the display of a Priority for Obujeku preparative video signal _V? Shah Doubideo signal (V ₄ or V _6). For example, II _0A = 1 when the priority of Obuji E click preparative video signal V ₂ is 1 0 0 Bruno. Is a Ichise emissions Bok, Shah Doubideo signal (V ₄ or V 6) is not displayed at all. That is, it can be said that the object video signal V 2 is opaque. H OA = Obujiweku Tobideo signal V 2 when the 0.5 is a semi-transparent, so that the sheet catcher Doubideo signal V ₄ are show through. That is, the object video signal V ₂ is displayed 50%, since Shah Doubideo signal (V ₄ or V ₆₎ is displayed 50%, Obujiweku preparative bi de O signals V ₂ and Shah Doubideo signal (V ₄ or V ₆ ) will be displayed.

Figure 2 1 shows a configuration of co Npaina 3 0 as a whole, and the first video signal combining unit 4 1 for calculating the synthesis ratio of the object video signal V _2, Shah Doubideo signal (V ₄ or V ₆₎ A second video signal synthesizing section 42 for calculating a synthesizing ratio; a key signal synthesizing section 43 for forming a key signal for the object video signal V ₂ and the shadow video signal (V ₄ or V ₆ ); and an object video signal V ₂ and shadow video signal (V ₄ or V _R) and co Nbaina 3 0 synthesized composite output unit 4 for outputting 4 and the ply o utility signal generating circuit 4 8 Metropolitan a is configured.

The first video signal synthesizing section 41 includes a “1” coefficient circuit 49, an “one-K ′ 4 (or one-K ′ ₆ )” coefficient circuit 50, a synthesizing circuit 45, and multiplication circuits 46, 4. Consists of seven.

 The combining circuit 45 calculates the following equation based on the first, second and third input data D 1, D 2 and D 3.

Output data D 4 is obtained by executing the operation of D 4 = D lx D 2 + (1−D 1) × D 3...-(85). Data D 1, D 2 and D 3 to be actually input and priority signal H _0A output from Buraio utility signal generating circuit, "1" from the coefficient circuit ώ "1" and output "1 - K _'4 ( Or 1 — Κ ' ₆ ) ”Since the“ 1 — K' ₄ (or 1 K ' ₆ ) ”output from the coefficient circuit 50 is input, the output data D 4 output from the combining circuit 45 Is {

Η θΑ + (1-Η ΟΑ) · (1-Κ '4 (or Κ' 6))} . Further, the output data D4 is supplied to multiplication circuits 46 and 47, and the multiplication circuit 46 receives the output data D4 and the object signal 信号₂ input to the compiler 30. output D 4 X kappa ₂ is obtained, and the output D 4 X kappa ₂ and object video signal V 2 is input to the multiplier 4 7 is al, an output D 4 X Κ ₂ XV ₂ is obtained. Therefore, the output data S 11 output from the first video signal combining unit 41 is

 S 1 1 {Η θΑ + (1 Η 0 A)

(1-K '(or K' ₆ ))} K 2-V 2

 · · · · (8 6).

The second video signal combining unit 4 3, "1 one H _0A" coefficient circuit 5 2 "1" and the coefficient circuit 5 3 "1 one K _2" coefficient circuit 5 4 and the same operation as (7 5) It comprises a synthesis circuit 51 to be executed, and multiplication circuits 152 and 153.

The input data D1, D2, and D3 input to the synthesis circuit 51 are respectively the " _111H0A " output from the " _111H0A " coefficient circuit ₅₁ and the " _111H0A " output from the "1" coefficient circuit. and "1" output, "1 - K _2" coefficient circuit 5 4 - "since 1 kappa _2" is output, output data Isseki D 4 output from the synthesizing circuit 5 1, {(1 - Η ΟΑ ) + Η ΟΑ · (1-Κ 2)}.

 Therefore, the combined output data S 1 2 of the second video signal combining unit 43 is

S 1 2 = {(1-Η ΟΑ) + Η ΟΑ · (1 — Κ ₂ )} Κ '4 (also

Κ '6)-V 4

It is expressed as

The key signal synthesizing unit 4 3, multiplication circuit 5 5 as "1 - (1 one kappa ₂₎ - (1 - kappa _'4 (or kappa' ₆₎₎ 'and an arithmetic circuit 5 6. The multiplication circuit 5 5 is expressed as “(1 一 ′ ₄ (or K ′ ₆ )) "From the coefficient circuit 5 0" (1 - K '4 (or kappa' ₆₎₎ "and output" (1 - kappa ₂₎ "from the coefficient circuit 5 4" (1 one kappa ₂₎ "output and is given, - outputs "(1 one _{Κ 2) · (1 K '} 4 ( or K'_6)).""1 one _{(1 - Κ 2) · (} 1 - K '4 ( or kappa'_6))" operation circuit 5 6 The "(1 _ kappa ₂₎ in · (1 - Κ '4

(Or K ' ₆ )))''output and the output is

 1-(1-Κ 2) · (1-Κ 'Α (or K' 6))

 ····· (8 8) The key signal 表. Is output as Also, the key signal 示 represented by this equation (78). Is output as an output signal from the combiner 30 to the next mixer 40.

Furthermore, the combined output unit 4 4, "1 one _{(1 - Κ 2) ♦ (} 1 - Κ '4 ( or K'6))" is converted to a fraction in a reverse circuit 5 7 an output, the multiplier circuit 5 8 The composite video signal V is obtained by multiplying the sum by the sum output of the adder circuit 59. From combiner 30 to mixer 4

Output to 0.

 The adder circuit 59 is provided with the combined output data S 11 and S 12 of the first and second video signal combiners 41 and 42, and thus the combined video signal V. Is

_{V o = ((H OAK 2} + (1- H OA) - (Ι- Κ '4 ( or _{K' 6))} Κ 2} · V + (1- Η ΟΑ) Κ / 4 ( or K _'6) + Η ΟΛ-(1- Κ ₂ ) Κ '4

(Or Κ ' ₆ )}-V 4 (or V ₆ ))

X 1 (1— Κ ₂ ) · (1-Κ ' ₄ (or K' ₆ ))}

 ····· (8 9)

Therefore, the mixed video signal V ' _miK ( _{V〃 in the} parallel light source mode) and the mixed key signal K' _mi K (K "in the parallel light source mode) actually output from the combiner 30 are (8 9) Expression — , (88).

The theoretically calculated equation (84) of the composite video output is the same as the composite video output V _{0A of} equation (89) obtained in the combiner 30 in FIG. 21, and thus the combiner 30 can be said to execute the synthesis theory of equations (75) to (84).

In the combiner 30 of FIG. 21, as shown in FIG. 20 (A), the images I and A are the depth information H. And H _s , the video signals V ₂ and V ₄ (or V ₆ ) representing the first and second images 〇 and A are key signals K ₂ and K ′ ₄ ( Or keying processing by Κ '6).

Depth information of image 〇. Is depth information H where image 0 is closer to screen 3A than image Α. In the range W 2 and W 4 which has, object video signal priority signal so as to output V ₂ H. _A (FIG. 20 (D)) is generated in the priority signal generation circuit 48.

In contrast, in the video A, the output range W 1 and Shah de Ubideo signal V ₄ in the range of W 3 have the depth information H _s closer image 0 to disk rie down surface 3 A (or V 6) The priority signal to perform is calculated as (11- _H0A ) (Fig. 20 (E)).

 Thus, the composite video signal V from the combiner 30. As Figure 2

0 (G), in the high priority ranges W2 and W4 of the object video 0, the object video is projected on the screen surface 3A and the shadow video A is displayed. In the ranges W1 and W3 in which the priority of the depth information Hs is high, the shadow image A is displayed on the screen surface 3A.

 Next, this priority signal will be described. In this embodiment, the priority signal generating circuit 48 is configured as shown in FIG.

, Obujeku preparative video signal V ₂ and Shah Doubideo signal is input No. V ₄ (or V ₆ ) depth information H. And H _s (FIG. 23 (A)) are received by a subtraction circuit 65, and the subtraction output S 21 (2H.1 H _s ) (FIG. 23 (B)) is applied to a multiplication circuit 66. gain signal S 2 2 supplied from the register 6 7 - after being multiplied by (G), the multiplication output S 2 3 (= (H. one H _s) G) (Fig. 2 3 (B)) is, Limiter

Given to 6-8.

Here, the gain signal S 22 (= G) changes the slope (that is, the rate of change of the composite depth ratio) of the subtraction output (g I. 1 H _s ) as shown in FIG. 23 (B). It has a function to do.

The limiter 68 limits the value of the multiplication output S 23 to a range of +0.5 0.5 as shown in FIG. 23 (C), and thus the difference of the depth information. When the level is +0.5 or -0.5 and the difference of the depth information is in the range of +0.5 to 0.5, the limiter output S24 is ( I. H _A ) Gives the value of G. The limiter output S 24 is added with the offset signal S 25 (= L.) Given from the offset register 70 in the adder circuit 69, and the added output is shown in FIG. As shown in (D), the priority signal generation circuit 48 outputs the priority signal H _0A whose value switches in the range of 1 to 0.

In the priority signal generating circuit 48 of FIG. 22, when the first and second images 〇 and て い る intersect, a range W 2 3 (see FIG. 3 (D)), the priority signal H _0A is the depth information H of the first and second video images A and A. And Hs, the slope varies according to the magnitude of the difference.

Therefore, in the range W 23 near the crossing point, the priority signal H _0A does not suddenly switch to 1 to 0 (or 0 to 1). Images 0 and A are seen through so that they overlap each other (Fig. 20 (G)). In other words, Since image 〇 and image A are mixed and displayed, image 0 and image A

In the vicinity where A intersects, the mixing ratio of the image I and the image I gradually changes, so that a display without discomfort can be performed.

 Thus, it is possible to obtain an effect on the special effect image in which the images 〇 and A change softly in the range W23 near the point where the first and second images 0 and A intersect. In addition, the width of the boundary area and the way of changing the image can be adjusted to the extent required by the operator by adjusting the value of the gain output S22 of the gain register 67.

 The combiner 22 is a depth information synthesis circuit with a NAM mix circuit configuration.

The composite depth information H is selected by selecting depth information of the images I and A (representing the position of the image closer to the screen surface 3A). As the ZH _S sent from the co Nbaina 2 2.

A mixed video signal (V ' _miは in point light source mode, V〃 _miは in parallel light source mode) and a mixed key signal (K' _{mix in} point light source mode) In the parallel light source mode, K〃 _mi )) is expressed by the above-mentioned equations (89) and (88), respectively.

4 0 and mixing video signals and mixtures keys signal is input, backgrinding La window down de signal V _'BK and is inputted finally output video signal V' OUT is output. Therefore, the output video signal V 'output from the mixer 40. _υτ is obtained by _converting equation ( ₈₈ ) and equation ( ₈₉ ) into equation (C) and equation

 It is obtained by substituting into (D). That is, in the point light source mode,

V OUT = Κ ' _{MI K} V _MIK + (1 — K') V BK

= {HOAK ₂ + (1-HOA) (I —K ' ₄ ) K ₂ } V ₂

+ {(1-HOA) K ^R ₄ + HOA (1-K ₂ ) K '4} V 2 + (1-Κ' ₄ ) (1 _K 2) V _ΒΚ The output video signal V〃 in the parallel light source mode. _UT is

V OUT = V _MI „+ (Ι — Κ ') V BK

= {H OAK 2 + (1-H OA) (1 — K '6) K 2} V ₂

+ {(1- H OA) K ' ₆ + H OA (1- K ₂ ) K ^/ 6} V 2

+ (1-Κ ' ₆ ) (1-K 2) V Β κ · · ·, (91).

Here, the real shadow generator 50 controls the gain of the output real shadow signal (K'4 in the point light source mode, '6' in the parallel light source mode). when Y%, when the point light source mode is K '4 = 0, when directional light mode is K' (9 0) become as _G = 0 and equation (9 1), respectively,

V. _UT = {H OA ₂ + (1-H OA) K ₂ } V 2

+ (1 K ₂ ) V BK

= K ₂ V ₂ + (1-1 K ₂ ) V BK

V "OUT = {H o AK 2 + (1-H OA) K 2} V ₂

 + (1-K 2) V BK

= K 2 V ₂ + (1-K ₂ ) V BK · · · · (93). In this case, the above equation (9 2) and at a point source mode (9 3), the output video signal during directional light mode is other than the portion of K ₂ V 2 is (1 - K ₂₎ V BK becomes If the gain of the real shadow signal is low, the knock ground signal V _BK is output at about 100%.

Further, the re Arusha dough generator 5 0, when the gain of the re Arusha synchronizing first signal is 1 0 0% at the point source mode is, K _'4 == 1, when directional light mode is kappa'. _{Since 6} = 1, equations (90) and (1) are

V OUT = H OAK ₂ V ₂ + {(1— H. _A ) + H _0A (1-K ₂ )} V 4

-H OAK ₂ V 2 + (1- H OAK ₂ ) V 4 V "ουτ = H OAK ₂ V 2+ {(1- H OA) + H OA (1- K ₂ )} V ₆

= H OAK ₂ V ₂ + (1-H OAK ₂ ) V 6 (95) In this case, both the point light source mode and the parallel light source mode are parts other than H _{0AK 2} V ₂ GMBH Doubideo signal (V ₄ or V ₆₎ are output 1 0 ϋ%. Normally, the shadow video signal is black on the tube and surface, and in this case it will be dark black and projected on the monitor screen.

 (10) Explanation of operation and effect of special effect device

 Next, the operation of the special effect device will be briefly described.

First, the operator operates a three-dimensional pointing device, keys, and the like provided on the control panel 5 to calculate a read address used in the special effect device of the present invention. Enter each parameter to be set. In here, the necessary parameter menu over data to the calculation of the read add-less, Pasupeku te I blanking value [rho Zeta, X _s axis of shadow coordinate system, the rotation of each of the Y _s axis and Z _s axis Angles Θ _x , 0 γ, Θ z), the origin of the shadow coordinate system (x _s , y so, z so), the type of light source indicating whether it is a parallel light source or a point light source, and the position of the light source is (Χ ι_, y L, z L) or (r, α, β) and the like. When the mode for automatically setting the origin of the shadow coordinate system is specified, data indicating which of the four corner points (a to d) of the source video signal is specified is provided. , Input from the control panel 5. First, an example in which a point light source is specified as a light source type will be described.

The CPU 8 receives these information from the control panel 5. The parameters are received and read in real time and reflected in the calculation of the address. Specifically, the CPU 8 monitors a change in a parameter supplied from the control panel 5 at a frame cycle, and calculates a read address based on the supplied parameter. Because of the parameters (bubc ^, b, ~ b 33 ') are calculated in the frame cycle. Therefore, these parameters can be changed in real time in the frame period according to the operation of the operator, and can be read out in real time according to the changed parameters. The address is calculated. The CPU 8 can also store these parameters in the RAM 7 for each frame as set values. At this stage, since the operator has not instructed the source video signal V 0 to perform 3D image conversion, the monitor screen 3 displays the source video signal V. Is displayed.

Next, the operator operates the three-dimensional pointing device provided on the control panel 5 to generate the source video signal V. Command 3D image conversion operation. When a three-dimensional image conversion is instructed by the operator, the CPU 8 generates a three-dimensional conversion matrix T specified by the operator. Is the parameter one _{another, r ~ r 33, i χ} ., The _z and s receiving co emission control panel 5 or al, reflects these parameters Ichita the operation of reading add-less in Li Alta Lee beam Let it. Specifically, the CPU 8 monitors a change in these parameters supplied from the control panel 5 at a frame cycle, and reads a read address based on the supplied parameters. Parameters for counting (bub aa, b, ~ b

₃₃ ') is calculated at the frame cycle. Next, CPU 8 receives took parameters rura ^ £ _xヽI gamma £ based on _ζ and s, each parameter b of the formula (8) three-dimensional transformation matrix T ₃₃ represented by ' , B 3 3 are calculated. Specifically, the equations (28) to (37) show that La main Ichita Γ,, by and this substituting the Γ Β ^ ί χ. £ v . £ z and s, Bruno, "lame Ichita b,, it is possible to find the ~ b _33. In addition, CPU 8 is , The received three-dimensional transformation matrix T.

Γ 33. ix. Iv. に 関 す る Parameters for τ, S and shadow coordinates' Θ θ 0 ν. Θ ζ. Χ s o. Y _s . , Z _s . , And light sources. Based on these parameters, X _L , Y, and Z _L , based on these parameters, the three-dimensional transformation matrix (T ₃₃ shadow ') —' each parameter b, calculates a ~ b ₃₃ '. The CPU 8 supplies the calculated parameters b,..., B ₃₃ to the read address generation circuit 14 of the object signal generator 10, and calculates the calculated parameters b. ₃₃ ′ is supplied to the read address generation circuit 24 of the shadow signal generation section 20.

The read address generation circuit 14 of the object signal generation unit 10 is configured as follows. ? 11 11 Receives the parameters 1 _{1 1} to 1) ₃₃ from 8 and receives the screen address (X _s , Y s) from the screen address generation circuit 9 and obtains the equation (13) and based on the (1 4), Obuji two click preparative signal read add-less (chi _Micromax, Upsilon ") to be generated in the frame period. the generated read address (chi _Micromax, Upsilon _Micromax) is Frame memory for video signal 12 and frame memory for key signal

13, respectively. As a result, the object video signal V ₂ is outputted from the frame memory 12, and the object key signal Κ ₂ is outputted from the frame memory 13.

On the other hand, the read address generation circuit 24 of the shadow signal generation section 20 receives the parameters b and -b ₃₃ ′ from the CPU 8 and receives the parameters from the screen address generation circuit 9. receiving a rie N'a dress (X s, Ys), based on the formula (5 8) and (5 9), the read address _{(X M ', Y M'} ) full rate-time cycle for shadow signal Generated by The generated read addresses (Χ _Μ ′, Υ „′) are supplied to the frame memory 22 for the video signal and the frame memory 23 for the key signal, respectively. As shown in Β, the frame memory 22 outputs the shadow video signal V ₄ , and the frame memory 23 outputs the shadow key signal Κ _4. Is forced. In Li Arusha dough generator 5 0 off Remume mode Li 2 3 Shah synchronizing first signal K 4 output from is input, ROM 5 0 based on Li Arushi catcher dough controller 5 0 3 force parameter Ichita H _s 4 are stored, and supplies the table 1 consisting of para menu over data H _s and the gain de one data read the gain data corresponding gain data, the gain control circuit 5 0 0. Thus, the gain control circuit 5 0 0 gives the gain according to the gain data supplied from the real shadow controller 5 0 3, with respect to Shah synchronizing first signal K ₄ input. The real shadow controller repeats the above processing.

 The shadow signal output from the gain control circuit 500 is horizontal

LPF 501 is then input to the vertical LPF 502. Li Arusha dough controller 5 0 3, based on the value of H _s used as depth information in the present embodiment to be input is stored in the R 〇 Micromax 5 0 4 of Riarusha dough controller 5 0 3, from a table consisting of H _s and off I filter coefficient data (see Fig. 1 9 C) issued reads the corresponding filter coefficient data and supplies the full I filter coefficient data in the horizontal LPF 5 0 1. The horizontal LPF 501 multiplies the input shadow key signal by the filter coefficient data from the real shadow controller 503 and outputs the result. Then, Li Arusha dough controller 5 0 3 is based on the H _s input, are stored in the R 0 M 5 0 4 of Li Arusha dough controller 5 ϋ 3, H s and off I filter coefficients The corresponding filter coefficient data is read from the table including the data (see FIG. 19C), and the read filter coefficient data is supplied to the vertical LPF 502. The vertical PF 502 multiplies the shadow signal from the horizontal LPF 501 by the filter coefficient data read from the real shadow controller 503, and finally the real shadow generator 5 and outputs 0 real Shah synchronization - signal K _'4 and to outputs. Therefore, as shown in FIG. 19A, the data of the shadow S near object 0 is Since the filter coefficient data multiplied by the object 近 く becomes larger, the contour becomes clear. Then, as the distance from the object increases, the value of the filter coefficient data to be multiplied becomes smaller, so that the data of the shadow S gradually becomes less contoured (blurred).

 Although the contour cannot be shown to be gradually blurred on the drawing, the shadow is gradually blurred as it moves away from the object.

Co Nbaina 3 0, when the object signal generation unit 1 0 receives the old Bujiwe click preparative video signal V ₂ and Obujiweku Toki first signal K ₂ together, shea catcher du signal generator 2 0 from Shah Doubideo signal V ₄及beauty It receives the real signal K ₄ ′ and generates a mixed video signal V _mix ′ and a mixed key signal K _{mi K} ′ based on the equation (a). The mixer 40 includes an externally supplied background video signal V _BK and a mixed video signal V _mi „output from the combiner 30.

And the mixed key signal K _mi> <', and generates an output video signal VO UT' based on the equation (c).

 Next, a case where a parallel light source is designated as the light source will be described.

CPU 8, first of all, co-down door from the control panel 5, parameters Ichita (^ X for each axis of the shadow coordinate, 0, Q 1, parameters related to the origin of theヽshadow coordinates (x _s., Ys o z _so ) and the parameters (r,, β) related to the collimated light source, and based on the operation state of the three-dimensional pointing device provided on the control sigma panel 5, the CPU 8 from Control This setup b Rupaneru 5, para bud 3D transformation matrix T. Ichita _{r 1 1 ~ r 33, £} X ί, £ z, receives the s. CPU 8 is supplied from copolyesters emissions collected by filtration Rupaneru 5 Changes in these parameters are monitored in real time in the frame cycle. And a parameter (!) For calculating the read address based on the supplied parameters! , ~! ^, B, -b ₃₃ ') are calculated at the frame period. Therefore, these parameters can be changed in real time in the frame cycle in accordance with the operation of the operator, and the read out data can be read out in real time in accordance with the changed parameters. The address is calculated.

Next, CPU 8 receives the parameters! ",! ~! · Calculates each parameter b ^ ~ 次 元₃₃ of the three-dimensional conversion matrix T ₃₃ — ¹ represented by equation (8) based on ₃₃ , ヽ. Specifically, equation (2 8) to formula (3 7), Bruno,. ra Ichita Τ · Η~ Γ _33, by substituting _"Ά Υ. iz and s, Bruno. Lame one! Can be obtained. Also, the CPU 8 receives the three-dimensional conversion matrix T. Para menu over data _{r,, ~ r 33, i} X. Si γ. SL z. S, and parameters _x about Shah de coordinates, 0 γ, Θ z, X s. , Y _s . , Z _s . And para menu over data r relates to a light source, receives the 9, based on these parameter menu over data, three-dimensional conversion matrix expressed in equation (7 1) (T ₃₃ Shadow〃) - 'each of parameters b, you calculating the ~ b ₃₃ ". CPU 8 is computed parameters Ichita b,, a ~ b _33, supplied to the talent Brzeg preparative signal generation unit 1 0 read § de Les generating circuit 1 4 and para menu over data b that computation, a ~ b ₃₃ ", and supplies the read address generating circuit 2 4 shadow signal generator 2 0.

Read § address generating circuit 1 4 objects signal generator 1 0, together with receive parameters b ,, ~ b ₃₃ from the CPU 8, the scan click rie N'a de-less generation circuit 9 crow click rie N'a Dress (X _s, Y s) receives, based on the formula (1 3) and (1 4), Obuji We click preparative signal read add-less (chi _Micromax, Upsilon _Micromax) to generate in frame period. The generated read add-less (Χ _Μ, Υ Μ) is off Remume mode for full Remume mode Li 1 2 and the key signal for bi de O signals Are supplied to the memory 13 respectively. As a result, an object video signal V ₂ is output from the frame memory 12, and an object video signal K ₂ is output from the frame memory 13.

- How, read § de-less generation circuit 2 4 shadow signal generator 2 0 para menu over data b from the CPU 8,, 〃 ~ b ₃ receive a ₃ ^ff, the Rutotomo, scan click rie N'a de Les generating circuit 9 crow click rie N'a de Les (X _s, Y _s) receive and formula (7 2) and (7 2) on the basis of the read for shadow signal add-less (XM " , generates the YM ') at full rate-time period. the generated read § de-less (X _M 〃, YM 〃) is full Remume motor Li for frame Note Li 2 2 and keys signal for a video signal 23, and as a result, as shown in FIG. 14B, the frame video 22 outputs the shadow video signal V _{6, and the} frame memory 23 outputs the shadow video signal K. ₈ is output.

In Li Arusha dough generator 5 0 Shah Douki signal K ₈ output from the full Remume mode Li 2 3 is input, R 0 Micromax 5 Riarusha dough control section 5 Y 3 is based on the parameters Ichita H _s The corresponding gain data is read from a table of parameters and gain data stored in 04, and the gain data is supplied to the gain control circuit 50 #. Therefore, the gain control circuit 500 gives a gain corresponding to the gain data supplied from the real shadow control section 503 to the input shadow key signal.

The shadow signal output from the gain control circuit 500 is input to the horizontal LPF 501 and then to the vertical LPF 502. Riarusha dough controller 5 0 3, based on the value of H _s which need use as depthwise information in this embodiment to be input, stored in the R 0 Micromax 5 0 4 of Li Arusha dough controller 5 0 3 are, from Ruteichi Bull such from H _s and the filter coefficient data (see Fig. 1 9 C) corresponding filter coefficient data Read and supply this filter coefficient data to the horizontal LPF 501. The horizontal LPF 501 multiplies the input shadow key signal by the filter coefficient data from the real shadow controller 503 and outputs the result. Then, Li Arusha de controller 5 0 3, based on the H _s input, re Arusha dough controller 5 0 3 R 〇 M 5 0 4, the stored, H s and off I filter The corresponding filter coefficient data is read from the table including the coefficient data (see Fig. 19C), and the read filter coefficient data is supplied to the vertical LPF 502. The vertical LPF 502 multiplies the shadow key signal from the horizontal LPF 501 by the filter coefficient data read from the real shadow controller 503, and finally the rear shadow generator 5 output as Li Arusha Douki signal K _'6 to output Y. Therefore, as shown in FIG. 19 (1), the contour of the shadow S data in the vicinity of the object becomes sharp because the filter coefficient data multiplied by the object becomes large. Then, as the distance from the object 0 increases, the value of the filter coefficient data to be multiplied becomes smaller, so that the data of the shadow S gradually becomes unsharpened (blurred).

 Although the contour cannot be shown as gradually blurring in the drawing, the shadow gradually blurs as it moves away from the object.

(/ ヽ ugu o

The combiner 30 receives the object video signal V ₂ and the object key signal K ₂ from the object signal generator 10, and also receives the shadow video signal V ₆ and the real signal from the shadow signal generator 20. It receives the dowkey signal K ₆ ′ and generates a mixed video signal V _mi „〃 and a mixed key signal K _{mi K} ″ based on equation (b). The mixer 40 outputs a background video signal V _BK supplied from the outside and a mixed video signal V _{mi K} * output from the combiner 30. And the mixed key signal K _{mi K} ′, and generates an output video signal V _{0 UT} ”based on the equation (d).

 (11) Effects of the present invention

 According to the present invention, a gain characteristic according to the depth information is controlled, and a filtering characteristic according to the depth information is controlled, so that a shadow for a more current object is generated, and an image of the shadow is generated. The object image and the package image are combined with each other, so with a simple configuration and simple processing, an object image that aims at a more realistic shadow according to the distance from the object image can be created at high speed. This has the effect that it can be added.

 In addition, the image special effect device for the object and the shadow can be used to create a more realistic shadow by a simple operation without forming a desired image by a separate operation. Can be synthesized.

 Furthermore, since a desired image is generated by the parameters of the operator, there is an effect that it is not necessary to spend an enormous amount of time to generate a desired image unlike computer graphics. Industrial applicability

 INDUSTRIAL APPLICABILITY The special effect device of the present invention can be used in a case where a special effect image is generated in a broadcast station image processing device.

Claims

Range of iif request  1. A special effect device for performing special effect processing on an object indicated by a video signal and a shadow of the object,  Gain control means for controlling the gain of the shadow image of the object,  Filtering means for filtering the shadow image,  Control means for controlling the gain of the gain control means according to the shadow depth information, and controlling the filtering characteristics of the filtering means according to the shadow depth information;  Synthesizing means for synthesizing an image of the shadow of the object, which is controlled and output by the control means, an image of the object, and an image serving as a background of the object;  Special effects device with. 2. The special effects device according to claim 2,  The control means has storage means in which gain characteristic data corresponding to the depth information and filtering characteristic data are stored,  The control means controls the gain of the gain control means in accordance with the gain characteristic data stored in the storage means, and controls the filtering characteristic of the filtering means in accordance with the filtering characteristic data stored in the storage means. Special effects equipment. 3. The special effects device according to claim 1,  The above-mentioned filtering means is a special effect device comprising a low-pass filter.  4. A special effect device that performs special effect processing on the input source video signal.  Object signal generating means for performing a first image conversion process on the source video signal to generate an object signal indicating a target image; A second image conversion process is performed on the source video signal. Shadow signal generating means for generating a shadow signal corresponding to the target image;  An object signal output from the object signal generation means and a shadow signal output from the shadow video signal generation unit are input, and the object signal, the shadow signal, and the source video are input. Synthesizing means for synthesizing a background signal corresponding to the signal and outputting an output video signal, wherein the shadow signal generating means includes: Gain control means for controlling gain, filtering means for filtering the shadow signal, depth information generating means for generating depth information corresponding to the shadow signal, and depth information from the depth information generating means. Control means for controlling the gain of the gain control means based on the depth information and controlling the filtering characteristics of the filtering means based on the depth information. Special effects apparatus having a door.  5. The special effects device according to claim 4,  The shadow signal input to the shadow signal generation means is input to the gain control means and output as a shadow signal whose gain is controlled, and the shadow signal whose gain is controlled is filtered by the filtering means. A special effect device that is input to the means, filtered, and output as a shadow signal from the filtering means.  6. The special effects device according to claim 4,  The above-mentioned filtering means is a special effect device comprising a one-pass filter. 7. In the special effects device described in the contract 4, The control means has storage means for storing gain characteristic data and filtering characteristic data according to the depth information from the depth information generating means, and stores the gain characteristic data and the filtering characteristic data. A special effect device for supplying the gain control means and the filtering means from the storage means according to the depth information, respectively.  8. The special effects device according to claim 4,  The object signal generation means includes a storage means for storing the input source video signal, and a read end address for reading the source video signal stored in the storage means in a predetermined unit. Read-out address generating means for generating the object video signal, wherein the first image conversion processing is such that the object video signal read from the memory means by the read-out address becomes a target image. Special effects equipment to be processed.  9. The special effects device according to claim 4,  The shadow video signal generation means includes a storage means for storing the input source video signal, and a read address for reading the source video signal stored in the storage means in a predetermined unit. Read-out address generating means for generating the video signal, wherein the second image conversion process is such that the shadow video signal read from the memory device by the read-out address corresponds to the target image. Special effect device processed to be output as a shadow signal 10. The synthesizing means receives the object signal from the object signal generating means and the shadow signal from the shadow signal generating means, and synthesizes the input signals to form a mixed signal. A special effect device comprising: a first synthesizing unit that outputs a video signal; and a second synthesizing unit that receives the mixing unit and the background signal and outputs an output video signal. 11. A special effect device for performing special effect processing on an input source video signal and a source key signal corresponding to the source video signal, When the source video signal and the source key signal are input, An object signal generating means for performing a first image conversion process to generate an object video signal indicating an object image and an object signal;  The source video signal and the sosky key signal are input A shadow signal generating means for performing a second image conversion process to generate a shadow video signal and a real shadow signal corresponding to the target image;  An object video signal and an object key signal output from the object signal generation means are input, and the shadow video signal and the above-mentioned real shadow key signal from the shadow signal generation means are input. And a synthesizing means for receiving a background signal corresponding to the source video signal, synthesizing the input signals to generate and output an output video signal, and ,  The shadow signal generation means includes: a shadow key signal generation means for generating a shadow key signal corresponding to the target image with respect to the input source key signal; and an output from the shadow key signal generation means. Gain control means for controlling the gain of the shadow key signal; filtering means for filtering the shadow key signal; depth information generating means for generating depth information corresponding to the shadow signal; and depth information generating means And a control means for controlling the gain of the gain control means based on the depth information and controlling the filtering characteristics of the filtering means based on the depth information.  12. The special effects device according to claim 12, The above-mentioned filtering means is a special effect device comprising a low-pass filter. 13. The special effects device according to claim 12,  The shadow key signal output from the shadow key signal generation means of the shadow signal generation means is output as a shadow key signal whose gain is controlled by the gain control means, and the shadow key whose gain is controlled. A special effect device in which one signal is input to the filtering means, and the wave characteristic is controlled and output as a real shadow key signal.  14. The special effects device according to claim 12,  The control means of the shadow signal generating section includes storage means for storing gain characteristic data and filtering characteristic data corresponding to the depth information from the depth information generating means, wherein the gain characteristic data and the filtering A special effect device for supplying characteristic data to the gain control means and the filtering means in accordance with the depth information, respectively.  15. The special effects device according to claim 12,  The synthesizing means receives the object video signal and the object signal from the object signal generating means, and outputs the shadow video signal and the real shadow signal from the shadow signal generating means. A first synthesizing means for generating a mixed video signal and a mixed key signal by synthesizing the input signals and a mixed video signal from the first synthesizing means; A second combining unit that receives the key signal and the background signal and generates the output video signal.  16. The special effects device according to claim 12, wherein The object signal generating means includes a first memory means for storing the source video signal, a second memory means for storing the source key signal, and the first memory means. The source videos stored in the memory means and the second memory means, respectively. A read address for reading the signal and the above-mentioned so-sky signal from the respective memory means is generated, and the read address is sent to the first memory means and the second memory means. A read address generating means for supplying an address, wherein the first image conversion processing is performed by the read address and the first memory means and the second memory itself are used. A special effect device in which the object video signal and the object signal read out from the memory device are output as the target image.  17. The special effects device according to claim 12,  The shadow signal generation means includes a third storage means for storing the source video signal, a fourth memory means for storing the source key signal, and a third memory means. And generating a read address for reading the source video signal and the source key signal stored in the memory means and the fourth memory means from the respective memory means. Read address generating means for supplying the read address to the third memory means and the fourth memory means, wherein the second image conversion processing comprises the read address And outputs the shadow video signal and the shadow signal read out from the third memory means and the fourth memory means, respectively, so that the shadow video signal is output for the above purpose. Shadow corresponding to image Special effects apparatus for processing output as a video signal.  18. A special effect method for performing special effect processing on an object indicated by an input video signal and a shadow of the object,  A gain step of controlling a gain of an image showing the shadow of the object, and a filtering step of filtering the image of the shadow, The gain of the gain step is controlled according to the depth information of the shadow, and the filtering step is controlled according to the depth information of the shadow. A control step of controlling the filtering characteristics of  A special effect including a combining step of combining the image of the shadow of the object output under the control of the control step, the image of the object, and the background image of the object to output an output video signal. Method.  19. In the special effects method described in $ 18  The control step includes controlling the gain in the gain step according to the gain characteristic data corresponding to the depth information stored in the storage unit, and converting the gain characteristic data into filtering characteristic data corresponding to the depth information stored in the storage unit. Thus, a special effect method for filtering the shadow image in the filtering step.  20. In the special effect method that performs special effect processing on the input source video signal,  An object signal generation step of performing a first image conversion process on the source video signal to generate an object signal indicating a target image;  A shadow signal generating step of performing a second image conversion process on the source video signal to generate a shadow signal corresponding to the target image;  A synthesizing step of receiving the object signal, the shadow signal, and the background signal corresponding to the source video signal, synthesizing the input signals, and outputting an output video signal; Has, The shadow signal step includes a gain control step of controlling a gain of the shadow signal corresponding to the source video signal, a filtering step of filtering the shadow signal, and a filtering step of filtering the shadow signal. A depth information generation step for generating depth information corresponding to the shadow signal; and adding the depth information from the depth information generation step to the depth information generation step. Controlling the gain of the gain control step based on the depth information and controlling the filtering characteristics of the filtering step based on the depth information.  21. an object signal generating step of performing a first image conversion process on an input source video signal to generate an object signal indicating a target image;  A shadow signal generation step of performing a second image conversion process on the source video signal to generate a shadow signal corresponding to the target image,  The object signal generated in the object signal generation step and the shadow signal generated in the shadow video signal generation step are input, and the object signal, the shadow signal, and the source video are input. Combining a background signal corresponding to the signal and outputting an output video signal.  The shadow signal generating step includes a gain control step of controlling a gain of the shadow signal corresponding to the source video signal, a filtering step of filtering the shadow signal, and a shadow signal. A depth information generating step for generating depth information corresponding to the depth information, and controlling the gain of the gain control step based on the depth information from the depth information generating step and filtering the filtering step based on the depth information. A special effect method having a control step of controlling wave characteristics.  The above filtering step is a special effect method performed by a low-pass filter.  22. The method of claim 21, wherein: The control step includes storing gain characteristic data and filtering characteristic data according to the depth information from the depth information generating step. A special effect method comprising: storing the gain characteristic data and the filtering characteristic data from the storing step to the gain control step and the filtering step, respectively, according to the depth information. 23. In the special effect method according to claim 21,  In the object signal generating step, a storage step for storing the input source video signal, and the source video signal stored in the storage step can be read out in a predetermined unit. A read address generating step of generating a read address, wherein the first image conversion processing is such that the object video signal read from the storage step by the read address becomes a target image. Special effects method to be processed.  24. The special effects method according to claim 2],  The shadow video signal generation step includes a storage step for storing the input source video signal, and a readout module for reading the source video signal stored in the storage step in a predetermined unit. A read address generating step of generating a dress, wherein the second image conversion process is a process in which the shadow video signal read from the memory means by the read address is a shadow corresponding to the target image. A special effect method that is processed to be output as a dough signal.  25. The special effect method according to claim 21, In the combining step, the object signal from the object signal generation step and the shadow signal from the shadow signal generation step are input, and the input signals are combined to synthesize a mixed signal. A special effect method comprising: a first synthesizing step of outputting an output video signal; and a second synthesizing step of receiving the mixed signal and the background signal and outputting an output video signal. 26. A special effect method for performing special effect processing on an input source video signal and a source key signal corresponding to the source video signal,  The source video signal and the source key signal are input, and an object signal generation step of performing a first image conversion process to generate an object video signal and an object key signal indicating a target image. ,  The source video signal and the source key signal are input, a second image conversion process is performed, and a shadow video signal corresponding to the target image and a shadow signal generation step for generating a real shadow key signal are provided. ,  The object video signal and the object key signal output from the object signal generation step are input, and the shadow video signal and the real shadow key signal from the shadow signal generation step are input. A synthesizing step for receiving a background signal corresponding to the source video signal, synthesizing the input signals to generate an output video signal, and outputting the synthesized video signal; The shadow signal generation step includes: a shadow key signal generation step of generating a shadow key signal corresponding to the target image with respect to the input source key signal; and an output from the shadow key signal generation step. A gain control step of controlling the gain of the shadow key signal to be performed, a filtering step of filtering the shadow key signal by 1, and a depth information generating step of generating depth information corresponding to the shadow signal. Controlling the gain of the gain control step based on the depth information from the depth information generation step and controlling the filtering characteristics of the filtering step based on the depth information. Method. 27. The method of claim 26, wherein:  The above-mentioned filtering step is a special effect method performed by a single pass filter.  28. The method of claim 26, wherein:  The shadow key signal output from the shadow key signal generation step in the shadow signal generation step is output as a shadow key signal whose gain has been controlled in the gain control step, and the shadow key signal whose gain has been controlled. A special effect method in which one signal is input to the above-mentioned filtering step, and the filtering characteristics are controlled and output as a real shadow key signal.  29. The method of claim 26, wherein:  The control step of the shadow signal generation step includes a storage step in which gain control data and filtering characteristic data corresponding to the depth information from the depth information generation step are stored. And a special effect method for supplying the gain characteristic data and the filtering characteristic data to the gain control step and the filtering step, respectively, according to the depth information.  30. The method of claim 26, wherein:  In the synthesizing step, the object video signal and the object signal from the object signal generating step are input, and the shadow video signal and the real shadow key signal from the shadow signal generating step are input. A first synthesizing step of synthesizing the input signals and generating a mixed video signal and a mixed key signal; and a mixed video signal and a mixed key from the first synthesizing step. A second combining step of receiving the signal and the background signal to generate the output video signal. 31. The method of claim 26, wherein: The object signal generating step includes a first memory step for storing the source video signal, a second memory step for storing the source key signal, and the first memory step. Generating a read address for reading the source video signal and the source key signal stored in the respective memory steps and the second memory step from the respective memory steps. A read address generating step of supplying the read address to the first memory step and the second memory step, wherein the first image conversion process According to the read address, the object video signal and the object key signal read from the first memory step and the second memory step, respectively, are read. Special effect method output as the target image. 32. The method of claim 26, wherein:  The shadow signal generation step includes a third storage step for storing the source video signal, a fourth memory step for storing the sowski signal, and a third memory step. And a read address for reading the source video signal and the source key signal stored in the memory step and the fourth memory step, respectively, from the respective memory steps. And a read address generating step of supplying the read address to the fourth memory step and the fourth memory step. The second image conversion process includes the read address. The shadow video signal and the shadow key signal read from the third memory step and the fourth memory step are output by the A special effect method in which a dough video signal is processed to be output as a shadow video signal corresponding to the target image.