JP2000057365A - Drawing device for three-dimensional graphics - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a drawing device for three-dimensional graphics capable of executing a drawing processing at high speed by using a parallel processing in a separate ALU(arithmetic and logic unit) system. SOLUTION: Polygon data is converted into a parameter required for drawing the polygon data by every scanning line, is stored in an intermediate information memory 301 by a polygon data converter 201, a parameter required for drawing by every one scanning line is calculated by using the parameter for drawing a polygon which is stored in the intermediate information memory 301, the contents of the intermediate information memory 301 are updated by a pre-processor 202, interpolation of a pixel value, implicit-surface elimination for drawing one scanning line are performed by using the parameter from the pre-processor 202 and the parallel processing in the separate ALU system by a pixel value interpolation device 203 and an implicit- surface eliminating device 204, a pixel value, a Z value are written in an intermediate drawing memory 302 in a form suitable for the parallel processing in the separate ALU system by a pixel drawing device 205, the pixel value for one scanning line which is stored in the intermediate drawing memory 302 is converted and written in an applicable position of the scanning line of a frame buffer after the drawing of one scanning line is completed by a pixel value converter 206.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a graphics drawing apparatus, and more particularly to a three-dimensional graphics drawing apparatus for displaying an image of a three-dimensional object on a two-dimensional screen by using a parallel processing of a divided ALU system.

[0002]

2. Description of the Related Art In recent years, as a microprocessor, a microprocessor having an operation instruction of a divided ALU (Arithmetic and Logic Unit: arithmetic logic unit) system is known. The operation instruction of the divided ALU system divides and uses one ALU into a plurality of ALUs, and performs the same processing on a plurality of data with one instruction.
on stream, Multiple Datastream) type parallel processing.

For example, as shown in FIG. 5, a 64-bit ALU is divided into four 16-bit ALUs, and divided A
The four 16-bit data are processed in parallel using the LU. Such a divided ALU parallel processing technology is described in the document (1) “MMX Technology Extension to the Intel
Architecture, IEEE MICRO, Vol.16, No.4, pp42-5
0, August 1996 ”, reference (2)“ The Design of the
Microarchitectureof UltraSPARC (TM) -I, Proceedin
gs of The IEEE, Vol. 83, No. 12, pp1653-1663, 19
December 1995 ”, Reference (3)“ Digital, MIPS Add Multim ”
edia Extensions, Microprocessor Report, Vol. 10,
No.15, pp24-28, November 1996 ”and reference (4)“ Su
bword Parallelism with MAX-2, IEEE MICRO, Vo
l.16, No. 4, pp 51-59, August 1996 ".
By using the parallel processing technique of the divided ALU system disclosed in these documents, it is possible to process signal processing data with high data parallelism at high speed.

[0004] Three-dimensional graphics processing frequently used in home-use TV game machines, personal computers and the like can also be sped up by parallel processing of the divided ALU system.

In the three-dimensional graphics processing, an object having a three-dimensional shape arranged in a three-dimensional space is converted into an image on a two-dimensional (flat) screen such as a television receiver or a CRT (Cathod Ray Tube) display device. indicate.

At this time, the three-dimensional shape is defined as a set of polygons (polygons) such as a triangle and a quadrangle, and each polygon is often drawn on a screen.

When rendering a polygon on a screen, coordinate conversion is performed to convert the vertex coordinates of the polygon into a coordinate system having the viewpoint as the origin, and illumination calculation for obtaining the amount of light emitted from the light source and reflected on the surface of the object to reach the viewpoint. Processing such as clipping for excluding a portion protruding from the screen, perspective projection for projecting from a three-dimensional space to a two-dimensional screen, shading for shading an object surface, and erasing a hidden surface for an object close to a viewpoint to hide an object far away are performed. . Further, in order to improve the quality of a drawing result, texture mapping for attaching a pattern to the surface of an object, alpha blending for expressing a translucent object, and the like may be performed.

In order to express a complicated scene or create a realistic image, it is effective to increase the number of polygons to be drawn. However, when the number of polygons increases, the processing amount required for drawing also increases. Therefore, it is necessary to increase the processing speed.

[0009] In the shading and hidden surface elimination processing for drawing the polygon projected on the two-dimensional screen to each pixel on the screen among the processing for drawing each polygon, each pixel or each pixel on the screen is displayed. Since the same processing is performed for each color component (the three components of red (R), green (G), and blue (B) are often used), it is effective to increase the speed by parallel processing of the divided ALU system.

A technique for speeding up three-dimensional graphics processing using the parallel processing of the divided ALU system is disclosed in reference (5), “3D Graphics Processing”.
Z-Buffer Using MMX (TM) Technology, Intel APP
LICATION NOTE, 1996 ”, Reference (6)“ Implementation of Glow Shading Algorithm Using MMX (TM) Instruction, Intel APPLICATION NOTE AP-5
38J, March 1996 ”, Reference (7)“ Using the MMX (TM) Instruction 2
Conversion from 4-bit true color data to 16-bit high color, Intel APPLICATION NOTE AP-5
53J, March 1996 ". In the techniques disclosed in these documents, glow shading is used as a shading method, and a Z buffer method is used as a hidden surface elimination method.

Glow shading is a shading method in which the color of each pixel inside a polygon is determined from the colors of the vertices of the polygon by bilinear interpolation. The Z-buffer method stores, for each pixel on the screen, a Z value representing the depth (distance from the viewpoint) of the polygon displayed on the pixel, separately from the data of the color displayed on the pixel. When drawing a polygon, the Z value of the polygon is obtained for each pixel in the area occupied by the polygon on the screen, and the Z value of the corresponding pixel is compared with the Z value already stored.
New polygon has smaller Z value (closer to viewpoint)
Only when the Z value and color data are updated to those of a new polygon, when all polygons are processed in this way, polygons farther from the viewpoint are sequentially overwritten by polygons closer to the viewpoint, and finally hidden surfaces An erased image is obtained.

FIG. 6 is a block diagram showing the configuration of a conventional three-dimensional graphics drawing apparatus using the technique disclosed in the above-mentioned literature and the like. The three-dimensional graphics drawing device includes a polygon data input device 100, a pre-processing device 501, a pixel value interpolating device 502, and a hidden surface removing device 50.
3, the pixel value conversion / drawing device 504, and the Z buffer 60
1, the frame buffer 602, the display device 400,
Consists of

FIG. 7 is a flowchart showing the processing of the conventional three-dimensional graphics drawing apparatus.

The operation of the conventional three-dimensional graphics drawing apparatus will be described below with reference to FIGS.

The frame buffer 602 is a memory for storing color data (hereinafter, referred to as “pixel value”) to be displayed on each pixel in the screen. The Z buffer 601 is a memory that stores a Z value of a polygon to be drawn on each pixel.

Before starting drawing, the frame buffer 6
The pixel value representing the background is stored in all the pixels 02, and the Z value of all the pixels in the Z buffer 601 is set to the maximum expressible Z value (step S201).

From the polygon data input device 100, after performing processing such as coordinate conversion, illumination calculation, clipping, and perspective projection for each polygon displayed on the screen,
The polygon data projected on the two-dimensional screen is input (step S202).

The input polygon data is, for example, 2
Two-dimensional coordinates X (i) and Y (i) of each vertex of the polygon projected on the three-dimensional screen, a Z value Z (i) indicating the depth of each vertex viewed from the viewpoint, and red, green, and blue of each vertex , R (i), G (i) and B (i), respectively. Here, i is a vertex number for distinguishing each vertex, and if the number of vertices of the polygon is N, i = 0, 1,. . . , N-1
It is.

The preprocessing device 501 calculates parameters for performing glow shading (hereinafter, referred to as "shading parameters") based on the polygon data input from the polygon data input device 100. The shading parameters include, for example, a slope DZDX of the Z value in the X axis direction, a slope DZDY of the Z value in the Y axis direction,
The inclination DRDX, DGDX of the color components R, G, B in the X-axis direction
And the gradient DR of the color components R, G, and B in the Y-axis direction.
DY, DGDY and DBDY. Further, the minimum value YMIN and the maximum value YMAX of the Y coordinate of the area occupied by the polygon on the screen are obtained, and the Y coordinate of each scanning line on the screen is Y.
For those having a value of MIN or more and YMAX or less, the X coordinate of the intersection of the polygon and the scanning line is calculated using the polygon data and the shading parameter, and an area (span) included in the polygon among the scanning lines is obtained.

Further, a parameter (hereinafter referred to as "span parameter") for performing pixel value interpolation and hidden surface elimination of each span is calculated. The span parameter includes, for example, the X coordinate XMIN at the left end of the span and the X coordinate X at the right end of the span.
MAX, Z value ZS at the left end of the span, and color components RS, GS, and BS at the left end of the span. The span parameter is, for example, sequentially from the one of the scan line span of the small Y coordinate, the pixel value interpolation device 502 and the hidden surface elimination device 5
03 (Step S203).

Pixel value interpolation device 502 and hidden surface elimination device 5
In step 03, pixel value interpolation and hidden surface removal processing are performed for each pixel included in each span of the polygon based on the shading parameters and span parameters supplied from the preprocessing device 501.

The pixel value interpolator 502 uses the color components RS, GS, BS at the left end of the span and the gradients DRDX, DGDX, DBDX of the color components in the X-axis direction to obtain the pixel value PR, Find PG and PB.

The pixel value at the leftmost pixel of the span is P
R = RS, PG = GS, PB = BS.

The pixel values of the pixel one pixel to the right of this pixel are: PR = RS + DRDX, PG = GS + DGDX, PB = BS + DBDX.

Similarly, j pixels from the left end of the span (j =
0, 1, 2,. . . ) The pixel values of the right pixel are: PR = RS + j × DRDX, PG = GS + j × DGDX, PB = BS + j × DBDX.

In order to obtain this pixel value, the inclination of the pixel value is sequentially added to the leftmost pixel value as shown in FIG.
Although FIG. 8 shows only the R component, the pixel values of each pixel are similarly obtained for the G and B components. Since this process is a process of performing the same operation (addition) for a large number of pixels, the speed can be increased by using the parallel processing of the divided ALU system.

In the above-mentioned document (6) “Implementation of glow shading algorithm using MMX (TM) instruction”, as shown in FIG.
A technique is described in which pixel values for four pixels are obtained by one instruction using four-parallel addition of the system.

In the four-parallel addition shown in FIG. 9, four 16-bit data a0, a stored in the 64-bit register 700 are used.
1, a2, a3 and four 16-bit data b0, b1, b2, b3 stored in the 64-bit register 701 are respectively combined and added in four parallels, and the result is 64b
This is the process of storing in the it register 702.

FIG. 10 is a schematic diagram showing how pixel value interpolation within a span is performed using 4-parallel addition. Here, only the R component is shown.

In order to perform the processing in four parallel, RS, DR
DX is expressed in a 16-bit fixed-point format.
10 is four numerical values of RS, RS, RS, and RS, and register 711 is 0, DRDX, 2 × DRDX, 3 × DR.
The four values of DX are stored in the register 714 as 4 × DRD.
X, 4 × DRDX, 4 × DRDX, and 4 × DRDX are stored, respectively.

Here, the register 710 is added to the register 711.
Are added in parallel, the pixel values of four pixels from the left end of the span are obtained in the register 712. Further, a register 713 (register 7) having the same contents as the register 712
13 may be the same register as the register 712) and the register 714 is added in four parallels, thereby obtaining the pixel values of the next four pixels. Hereinafter, pixel values for four pixels are sequentially obtained by sequentially adding four registers 714 in parallel (step S204).

The hidden surface elimination device 503 obtains the Z value PZ of each pixel in the span using the Z value ZS at the left end of the span and the gradient DZDX of the Z value in the X-axis direction. By comparing with the Z value recorded at the pixel position, a pixel whose PZ is smaller than the Z value stored in the Z buffer is determined, and the Z buffer at that pixel position is updated to PZ.

The Z value of each pixel in the span is obtained in the same manner as the pixel value interpolation by the pixel value interpolation device 502. Since the comparison of the Z values is a process of performing the same operation (comparison) for a large number of pixels, the speed can be increased by using the parallel processing of the divided ALU system.

The above document (5) “3D Z-Buffer”
Using MMX (TM) Technology "
Describes a technique in which a Z value comparison for four pixels is performed by one instruction using a four-parallel comparison of a divided ALU system as shown in FIG.

The four-parallel comparison shown in FIG. 11 is based on the four 16-bit data a0, a stored in the 64-bit register 720.
1, a2, a3 and four 16-bit data b0, b1, b2, b3 stored in the 64-bit register 721 are respectively combined to perform four-parallel magnitude comparison.
When n> bn (n = 0, 1, 2, 3) holds,
0xFFFF (“0x” indicates that the numerical value is a hexadecimal number) is set to 64 bits when an> bn (n = 0, 1, 2, 3) is not satisfied.
This is a process of storing in the register 722. The contents of the register 722 in FIG. 11 represent an example of the result.

FIG. 12 is a schematic diagram for explaining the manner in which the Z value is compared using 4-parallel comparison. In order to perform the processing in four parallels, the Z value is represented by a 16-bit fixed point.
Z at each pixel in the span determined in the same manner as pixel value interpolation
The value is stored in the register 731, the Z value read from the corresponding pixel position in the Z buffer 601 is stored in the register 730, the contents are compared in four parallels, and the result is stored in the register 731.
2 is stored.

The contents of the register 732 are set to 0xFFFF for a pixel whose Z value and pixel value need to be updated, and set to 0x0000 for a pixel which does not need to be updated. The Z value of the position is updated.

More specifically, for example, as shown in FIG. 13, the Z value of the pixel position at which the Z value is updated is stored in the register 7 by the 4-parallel AND of the register 732 and the register 731.
33, which is the negation of register 732
The Z value of the pixel position not to be updated is extracted to the register 735 by the four parallel logical product of the register 34 and the register 730, and the four parallel logical sum of the register 733 and the register 735 is obtained.
The updated Z value is obtained in the register 736. When processing the beginning or end of a span, it may be necessary to perform drawing of less than four pixels. In this case, for example, the Z value comparison result (register 732) of the hidden surface erasing device 503 may be used. 0x0000 at positions corresponding to unnecessary pixels
Is set to avoid unnecessary drawing of pixels (step S205 in FIG. 7).

The pixel value conversion / rendering device 504 is based on the pixel values within the span determined by the pixel value interpolator 502 and the Z value comparison result at each pixel position determined by the hidden surface elimination device 503. , The pixel value at the corresponding pixel position in the frame buffer 602 is updated. Here, the pixel value interpolation device 50
2, the frame buffer 602
Is converted to be stored in.

As shown in FIG. 14, the pixel value obtained by the pixel value interpolation device 502 is 16 for each color component PR, PG, and PB.
It is represented by a bit fixed point, and stored in a 64-bit register every four pixels.

On the other hand, in the frame buffer 602, for example, as shown in FIG. 15, the color components R, G, and B are continuously arranged for each pixel, and each color component is 8 bits (24 bits full color) or 5 bits. (16-bit high color) in many cases. Therefore, in order to update the pixel value of the frame buffer,
Conversion of the pixel value expression is performed (step S20 in FIG. 7).
6).

FIG. 16 is a schematic diagram showing how the pixel value expression is converted. For this pixel value conversion, the number of bits of each color component of the pixel value is converted (for example, from 16 bits to 8 bits).
t, 16bit to 5bit, etc.),
In addition, it is necessary to rearrange and position the pixel values.

Reference (6), "Implementation of Glow Shading Algorithm Using MMX (TM) Instruction", Reference (7), "From 24 Bit True Color Data Using MMX (TM) Instruction" 16
"Conversion to bit high color" discloses a pixel value conversion technique using parallel processing of a divided ALU system. Specifically, for example, the conversion into 24-bit full color is performed by using the interleaving process of the divided ALU system as shown in FIG. 17 and performing the conversion process as shown in FIG.

The interleave processing shown in FIG.
Of the four 16-bit data a3, a2, a1, a0 stored in the bit register 752, a1 and a0, 64
Of the four 16-bit data b3, b2, b1, and b0 stored in the bit register 751, b1 and b0 are interleaved, and b1, a are stored in the 64-bit register 753.
This is processing for storing in the format of 1, b0, b0.

In the conversion processing shown in FIG. 18, the color components of four pixels stored in the 64-bit registers 760, 761, and 762 in the format of 16 bits × 4 are subjected to the 4-parallel shift processing or the 4-parallel lower 8-bit clear. To extract the upper 8 bits of each component (registers 763 and 763).
64, 765), which are rearranged by using OR and interleave. However, since the conversion results obtained in the registers 767 and 768 include 0 in addition to the R, G, and B pixel values, the frame buffer 602
When writing to the frame buffer, it is necessary to further perform positioning in order to conform to the structure of the frame buffer shown in FIG.

Steps S204 to S2 in FIG.
At 06, drawing of four pixels of one polygon and one span is executed.

Subsequently, in order to draw all the pixels in one span, the end of the span is determined. Specifically, for example, four pixels are drawn from the pixel at the X coordinate XMIN at the left end of the span, and it is determined whether the drawing process has reached the pixel at the X coordinate XMAX at the right end of the span. If the pixel has not reached the right end, the process returns to step S204. The next four pixels are drawn by moving to the right by four pixels (step S207 in FIG. 7).

Steps S204 to S207
Then, drawing of one span of one polygon is executed.

Subsequently, in order to draw all spans in one polygon, the end of the polygon is determined. Specifically, it is determined whether or not the drawing process has reached the span of the Y coordinate YMAX of the polygon. If not, the process returns to step S203 in FIG. 7 to draw a span of one larger Y coordinate (step S208).

Steps S204 to S2 in FIG.
At 08, rendering of one polygon is executed. continue,
It is determined whether the rendering of all polygons has been completed. If the rendering has not been completed, the process returns to step S202 to input the next polygon data and render the next polygon (step S209 in FIG. 7).

When the rendering of all the polygons is completed, the display device 400 reads the contents of the frame buffer 602 and
The image is displayed on a television receiver or a CRT display device (step S210 in FIG. 7).

[0052]

However, in the conventional three-dimensional graphics rendering device disclosed in the above-mentioned documents, the pixel value conversion / rendering device 504 converts the pixel value stored in the frame buffer 602 into a pixel value. In order to directly update, it is necessary to convert the pixel value obtained by the pixel value interpolation device 502 according to the structure of the frame buffer 602.
There is a problem.

In the parallel processing of the divided ALU system, for example, 8
Processing of data in a fixed format such as bit × 8, 16 bit × 4, 32 bit × 2, etc. can be executed at high speed by parallel processing. However, a fixed process such as 5 bit × 3 in the case of a 16-bit high color pixel value is performed. If you want to handle data that does not match the format, you need to perform data format conversion using shift processing and mask processing, etc.
The processing speed is reduced due to the overhead.

In the Z-buffer method, pixel values stored in the frame buffer are sequentially overwritten by polygons closer to the viewpoint. Therefore, the pixel value conversion / rendering device 504 executes the above-described pixel value conversion processing also on the pixel value of a pixel which is overwritten by another polygon and is not actually displayed on the screen.

Such pixel value conversion processing of a pixel that is not actually displayed occurs many times, particularly when many objects are arranged in a three-dimensional space, and the overhead of the conversion processing causes In addition, the processing performance of the three-dimensional graphics drawing apparatus is reduced.

Accordingly, the present invention has been made in view of the above problems, and an object of the present invention is to provide a three-dimensional graphics drawing apparatus which can execute drawing processing at high speed by utilizing the parallel processing of the divided ALU system. Is to do.

[0057]

In order to achieve the above object, the present invention provides a three-dimensional graphics rendering apparatus for displaying an image of a three-dimensional object on a two-dimensional screen using a parallel processing of a divided ALU system. Polygon data conversion means for converting input polygon data into parameters necessary for drawing for each scanning line and storing the converted data in intermediate information storage means; and for polygon drawing stored in the intermediate information storage means. Preprocessing means for obtaining parameters necessary for rendering for each scanning line and updating the contents of the intermediate information storage means using the above parameters, and parallel processing of the divided ALU system using the parameters supplied from the preprocessing means. A pixel value interpolation unit and a hidden surface elimination unit for respectively performing pixel value interpolation and hidden surface elimination for drawing one scanning line by using
A pixel drawing unit that writes pixel values and Z values in the intermediate drawing storage unit in a format suitable for the U-type parallel processing, and one scanning line stored in the intermediate drawing storage unit after drawing of one or more scanning lines is completed And a pixel value conversion unit for converting the pixel value for each minute into a corresponding scanning line position of the frame storage unit.

[0058]

Embodiments of the present invention will be described. According to the present invention, in a preferred embodiment, the polygon data conversion device (201) converts input polygon data into parameters necessary for drawing each scanning line, and stores the parameters in the intermediate information memory (301). The preprocessing device (202) uses the parameters for polygon drawing stored in the intermediate information memory (301) to obtain parameters necessary for drawing for each scanning line and updates the contents of the intermediate information memory (301). , Pixel value interpolation device (20
3) The hidden surface erasing device (204) is a pre-processing device (20).
Using the parameters supplied from 2), pixel value interpolation and hidden surface elimination for drawing one scan line are performed by using the parallel processing of the divided ALU system.
Intermediate drawing memory (3) in a format suitable for LU-type parallel processing
02) is written in the pixel value conversion device (2).
In step 06), after drawing of one or a plurality of scanning lines is completed, the pixel values for one scanning line stored in the intermediate drawing memory (302) are converted and written to the corresponding scanning line position in the frame buffer.

More specifically, a polygon data conversion means (201) for converting input polygon data into parameters necessary for drawing each scanning line, and parameters obtained by the polygon data conversion means are displayed on a screen. Intermediate information storage means (30) for storing in association with each scanning line
1) and a preprocessing means for obtaining parameters necessary for drawing for each scanning line using the polygon drawing parameters stored in the intermediate information storage means (301) and updating the contents of the intermediate information storage means (1). 202), an intermediate drawing storage unit (302) for storing a pixel value and a Z value for each pixel of one or a plurality of scanning lines, and a polygon among the scanning lines using parameters obtained by the preprocessing unit (202). Value interpolating means for obtaining the pixel value of the pixel in the area occupied by the
Using the parameters obtained by the preprocessing means, the Z value of the pixel in the area occupied by the polygon in the scanning line is obtained, and the Z value stored at the position of the corresponding pixel in the intermediate drawing storage means (302) is obtained. Hidden surface erasing means (204) for performing hidden surface erasing processing using the pixel value obtained by the pixel value interpolating means (203) and the hidden surface erasing processing result of the hidden surface erasing means (204), and storing the intermediate drawing A pixel drawing means (205) for drawing pixel values in the means (302), a frame storage means (303) for storing pixel values of all pixels of one screen, and one or more scans from the intermediate drawing storage means (302) The pixel value of each pixel of the line is read out, the pixel value is converted into a format suitable for recording in the frame storage unit (303), and the pixel value conversion unit (2) is written to the corresponding scanning line position of the frame storage unit.
06).

The pixel value interpolation means (203)
The pixel value interpolation is performed using parallel processing of the divided ALU method, the hidden surface removing means (204) performs hidden surface removal using the parallel processing of the divided ALU method, and the pixel drawing means (205) performs the divided AL processing.
The intermediate drawing storage means (3) in a format suitable for U-type parallel processing
02), and a pixel value conversion means (206)
Reads out a pixel value in a format suitable for the parallel processing of the divided ALU system stored in the intermediate drawing storage means (302) after drawing of one or a plurality of scanning lines, and records it in the frame storage means (303) And writes it to the corresponding scanning line position.

In the embodiment of the present invention, polygon data conversion means (201), preprocessing means (202), pixel value interpolation means (203), hidden surface elimination means (204), pixel drawing means (205), The pixel value conversion means (206)
The function may be realized by a program executed by the data processing device (200). in this case,,
Polygon input means (100), data processing device (20
0), in a three-dimensional graphics rendering apparatus including a storage device (300) and a display means (400), the present invention is realized by loading and executing a program from a recording medium on which the program is recorded to a main memory of a data processing device. Will be implemented.

According to the embodiment of the present invention, polygon drawing is performed for each scanning line, and pixel value interpolation, hidden surface removal, and drawing within one or a plurality of scanning lines are performed using an intermediate drawing memory. In the intermediate drawing memory, pixel values and Z values are stored in a format suitable for the parallel processing of the divided ALU system, and after the drawing of one or a plurality of scanning lines is completed, the pixels stored in the intermediate drawing memory by the pixel value conversion device are stored. By converting the value and writing the converted value to the frame buffer, the amount of pixel value conversion processing which becomes an overhead during the drawing process using the parallel processing of the divided ALU system can be reduced, and the drawing process can be executed at high speed.

This is because, in the conventional three-dimensional graphics rendering apparatus, the pixel value conversion processing is also performed on the pixel values of the pixels that are overwritten by other polygons and are not actually displayed on the screen. According to the present invention, it is only necessary to perform one pixel value conversion process for each pixel in the screen, which is the minimum number of times. Hereinafter, the present invention will be described in detail with reference to examples.

[0064]

FIG. 1 is a block diagram showing the configuration of an embodiment of a three-dimensional graphics drawing apparatus according to the present invention. Referring to FIG. 1, an embodiment of a three-dimensional graphics rendering apparatus according to the present invention includes a polygon data input apparatus 100, a polygon data conversion apparatus 201, a preprocessing apparatus 202, a pixel value interpolation apparatus 203, a hidden surface It includes an erasing device 204, a pixel drawing device 205, a pixel value conversion device 206, an intermediate information memory 301, an intermediate drawing memory 302, a frame buffer 303, and a display device 400.

The intermediate information memory 301 is a memory for storing parameters necessary for rendering a polygon for each scanning line in association with each scanning line. The intermediate drawing memory 302 is a memory that stores a pixel value and a Z value for each pixel of one scanning line. FIG. 3 is a schematic diagram showing the structure of the intermediate drawing memory. As shown in FIG. 14, for example, as shown in FIG. 14, a pixel value of four pixels is 64 bits in a 16-bit × 4 format, as shown in FIG.
It is stored in the it register, and each color component R, G,
B is stored in another register. Intermediate drawing memory 3
02 has a structure as shown in FIG. 3, so that the obtained pixel value can be directly written without performing the pixel value conversion processing that is a processing overhead. In addition, since the Z value is similarly obtained in units of four pixels, the Z value is also stored in the intermediate drawing memory 302 continuously to the pixel value of the corresponding pixel as shown in FIG.

The frame buffer 303 is a memory for storing a pixel value of each pixel in the screen.

FIG. 2 is a flowchart showing the processing flow of one embodiment of the three-dimensional graphics drawing apparatus of the present invention.

The operation of one embodiment of the three-dimensional graphics drawing apparatus of the present invention will be described below with reference to FIGS. Here, for the sake of simplicity, a case where drawing is performed for each scanning line will be described as an example. Further, a case where pixel value interpolation and hidden surface elimination by parallel processing of the divided ALU method are performed in the same manner as described in the conventional method will be described.

Before starting drawing, the intermediate drawing memory 30
The pixel value representing the background is stored as the pixel value of all the pixels 2 and the maximum expressible Z value is stored as the Z value of all the pixels (step S101).

From the polygon data input device 100, processes such as coordinate conversion, illumination calculation, clipping, and perspective projection were performed for each polygon to be displayed on the screen, as in the case of the conventional three-dimensional graphics drawing device. After,
The polygon data projected on the two-dimensional screen is input (step S102).

The polygon data converter 201 performs parameters for performing glow shading based on the polygon data input from the polygon data input device 100 (hereinafter referred to as "shading parameters").
Is calculated. Then, the obtained shading parameter and span parameter are stored in Y of the intermediate information memory 301.
The coordinates are stored in a position corresponding to the scanning line of YMIN (step S103).

In order to perform the polygon data conversion for all the input polygons, it is determined whether the polygon data conversion for all the polygons has been completed. If not, the flow returns to step S102 to input the next polygon data. (Step S104).

When the conversion of all the polygon data is completed, the preprocessing device 202 reads out the parameters for drawing the polygon corresponding to the scanning line of a certain Y coordinate SY stored in the intermediate information memory 301, The parameters are supplied to the pixel value interpolation device 203 and the hidden surface removal device 204. Specifically, for example, first, SY = 0 is set, a parameter for drawing a polygon whose Y coordinate is associated with a scanning line of SY is supplied, and a polygon drawing whose Y coordinate is associated with a scanning line of SY is provided. When the drawing using the parameters for is completed, SY is incremented by 1 and the same processing is performed.

When supplying the parameters, the intermediate information memory 301 is updated. Specifically, when the maximum value YMAX of the Y coordinate of the polygon being processed is greater than or equal to SY + 1, the polygon being processed has a span even on the scanning line whose Y coordinate is SY + 1. The span parameter is obtained and the intermediate information memory 30
No. 1 is stored at a position corresponding to the scanning line whose Y coordinate is SY + 1 (step 105).

The pixel value interpolation device 203 performs the same processing as the pixel value interpolation device 502 in the conventional three-dimensional graphics drawing device. That is, based on the parameters received from the preprocessing unit 202, the pixel value of each pixel in the span is obtained by the parallel processing of the divided ALU system (step 106).

The hidden surface erasing device 204 performs the same processing as the hidden surface erasing device 503 in the conventional three-dimensional graphics drawing device. However, the intermediate drawing memory 302 is used for storing the Z value. That is, based on the parameters received from the preprocessing unit 202, the Z value of each pixel in the span is obtained by parallel processing of the divided ALU system, and compared with the Z value stored in the corresponding pixel position of the intermediate drawing memory 302. Then, the determination of the pixel whose pixel value is to be updated and the update of the Z value are performed (step S107).

The pixel drawing device 205 includes the pixel value interpolation device 2
03 and the pixel value in the span obtained in step 03
The pixel value of the corresponding pixel position in the intermediate drawing memory 302 is updated based on the result of the Z value comparison at each pixel position obtained in 04. Updating of the pixel values for four pixels can be performed by the same processing as updating the Z value in the conventional three-dimensional graphics rendering apparatus.

Here, the format of the updated pixel value stored in the intermediate drawing memory 302 is the same as the format of the pixel value obtained as a result of the parallel processing of the divided ALU system, as shown in FIG. Therefore, it is not necessary to perform the pixel value conversion processing required in step S206 in the conventional three-dimensional graphics drawing apparatus (step S1).
08).

The steps S106 to S1 in FIG.
At 08, drawing of four pixels of one scanning line and one span is executed.

Subsequently, in order to draw all the pixels in one span, the end of the span is determined. Specifically, for example, four pixels are drawn from the pixel at the X coordinate XMIN at the left end of the span, and it is determined whether the drawing processing has reached the pixel at the X coordinate XMAX at the right end of the span. If the pixel has not reached the right end, the process returns to step S104. The image is moved right by four pixels to draw the next four pixels (step S109). This is the conventional 3
Step S in FIG. 7 in the three-dimensional graphics drawing apparatus
This is the same processing as 207.

Steps S106 to S1 in FIG.
In the process 09, drawing of one scanning line and one span is executed.

Subsequently, in order to draw all spans in one scanning line, the end of the scanning line is determined. In particular,
For example, it is determined whether or not there is a parameter in the intermediate information memory for drawing a polygon associated with the scanning line of the Y coordinate SY, for which the span drawing has not been completed yet.
If the drawing of all spans in one scanning line has not been completed, the process returns to step S105, and the next span in that scanning line is drawn (step S110).

Step S 105 to step S 1 in FIG.
In the process of 10, rendering of all pixels of one scanning line in the intermediate rendering memory 302 is executed.

Subsequently, the pixel values of all the pixels in the scanning line are converted into a format suitable for being stored in the frame buffer 303, and written to the corresponding scanning line position in the frame buffer 303. The conversion of the pixel value is the same as the pixel value conversion performed by the pixel value conversion / drawing device 504 in the conventional three-dimensional graphics drawing device (step S11).
1).

Step S 105 to step S 1 in FIG.
In the process of 11, rendering of all pixels of one scanning line on the frame buffer 303 is executed.

Subsequently, in order to draw all the scanning lines in the screen, it is determined whether or not the drawing of all the scanning lines has been completed. If not, the flow returns to step S105 to draw the next scanning line. (Step S112).

When the drawing of all the scanning lines is completed, the display device 4
00 reads the contents of the frame buffer 303 and displays them on a television receiver or a CRT display device (step S113).

FIG. 4 shows the intermediate drawing memory 30 shown in FIG.
A different intermediate drawing memory 302 with a slight modification of the configuration of FIG.
An example of the configuration will be described.

In the above-described embodiment using the intermediate drawing memory 302 having the configuration shown in FIG. 3, the pixel drawing device 205 does not need to convert pixel values when writing pixel values into the intermediate drawing memory 302, and can perform drawing processing at high speed. Although it can be executed, in the intermediate drawing memory 302, it takes 16 to store one color component of one pixel.
In order to use a storage area of bits, the intermediate drawing memory 302
In some cases, a problem arises in that the amount of memory required for the operation becomes large.

Therefore, as a second embodiment of the present invention, an example using the intermediate drawing memory 302 having the configuration shown in FIG. 4 will be described. In this embodiment, 30 in the intermediate drawing memory
The storage capacity of 8 bits is used to store one color component of one pixel in 2.

Normally, when a pixel value is stored in the frame buffer 303, 8 b
This is because it may be less than it. In this embodiment, when the pixel drawing unit 205 writes pixel values in the intermediate drawing memory 302, it is necessary to convert the number of bits from 16 bits × 4 to 8 bits × 4. Since it is not necessary to rearrange and align the data required by the pixel value conversion / rendering device 504, the data can be easily processed as compared with the conversion process in the conventional three-dimensional graphics rendering device.

In this embodiment, the amount of memory required for the intermediate drawing memory 302 can be greatly reduced without greatly increasing the processing amount at the time of writing to the intermediate drawing memory 302.

In the above embodiment, for the sake of simplicity, the case where the intermediate drawing memory 302 stores the pixel value and the Z value of each pixel of one scanning line and performs the drawing process for each scanning line will be described. However, the intermediate drawing memory 302 may store the pixel value and the Z value of each pixel of a plurality of scanning lines, and perform the drawing process in units of a plurality of scanning lines.

In the above-described embodiment, the case where R, G, and B, which are the color components of the pixel values, are used to draw the polygon has been described. However, the present invention is applicable only when R, G, and B are used. The present invention is not limited to this. For example, the present invention can be applied to a case where an index value of a lookup table of pixel values is used.

In the above-described embodiment, a case has been described where four color components of pixel values of four pixels are calculated in parallel in pixel value interpolation using parallel processing of the divided ALU system. The present invention is not limited to this type of parallel processing, and is also applicable to, for example, a case where four components of one color component R, G, B and four alpha values used for alpha blending are processed in four parallel processes. be able to. In this case, the structure of the pixel values stored in the intermediate drawing memory 302 is a structure suitable for storing the pixel values obtained by the pixel value interpolation.

In the above-described embodiment, the case where the 16-bit × 4 4-parallel processing is performed by the divided ALU system is described. However, the present invention is not limited to the 16-bit × 4 format parallel processing. , 32 bits × 2, and the like.

The three-dimensional graphics drawing apparatus of the present invention can be realized by using a general computer system without using a dedicated system. For example, a three-dimensional graphics rendering apparatus that executes the above-described processing is configured by installing the program for realizing the above-described operation in a computer from a medium (a floppy disk, a CD-ROM, or the like) storing the program. be able to.

[0098] The medium for supplying the program to the computer may be a communication medium (a medium that temporarily and fluidly stores the program, such as a communication line, a communication network, or a communication system). For example, the program may be posted on a bulletin board (BBS) of a communication network and distributed via the network. Then, by starting this program and executing it in the same manner as other application programs under the control of the OS, the above-described processing can be executed.

[0099]

As described above, according to the present invention,
Polygon rendering is performed for each scanning line, and pixel value interpolation, hidden surface removal, and rendering within one or more scanning lines are performed using an intermediate rendering memory.
The pixel value and the Z value are stored in a format suitable for the parallel processing of the system, and after the drawing of one or a plurality of scanning lines is completed, the pixel value stored in the intermediate drawing memory is converted by the pixel value conversion device, By writing to the buffer, the amount of processing for pixel value conversion, which is an overhead during drawing processing using the parallel processing of the divided ALU system, can be reduced, and the drawing processing can be executed at high speed. To play.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of an embodiment of a three-dimensional graphics drawing apparatus according to the present invention.

FIG. 2 is a flowchart illustrating a process of an embodiment of the three-dimensional graphics drawing apparatus according to the present invention.

FIG. 3 is a schematic diagram showing an example of a configuration of an intermediate drawing memory in one embodiment of the three-dimensional graphics drawing device of the present invention.

FIG. 4 is a schematic diagram showing another configuration example of an intermediate drawing memory in one embodiment of the three-dimensional graphics drawing apparatus of the present invention.

FIG. 5 is a schematic diagram illustrating parallel processing of a divided ALU system.

FIG. 6 is a block diagram illustrating a configuration of a conventional three-dimensional graphics drawing apparatus.

FIG. 7 is a flowchart illustrating processing of a conventional three-dimensional graphics drawing apparatus.

FIG. 8 is a schematic diagram illustrating a state of pixel value interpolation.

FIG. 9 is a schematic diagram illustrating 4-parallel addition of the divided ALU system.

FIG. 10 is a schematic diagram illustrating pixel value interpolation using 4-parallel addition.

FIG. 11 is a schematic diagram illustrating four-parallel comparison of the divided ALU system.

FIG. 12 is a schematic diagram illustrating a Z-value comparison using 4-parallel comparison.

FIG. 13 is a schematic diagram illustrating Z value updating using four parallel logical operations.

FIG. 14 is a schematic diagram showing a structure of a pixel value obtained by a pixel value interpolation device.

FIG. 15 is a schematic diagram illustrating an example of the structure of a frame buffer.

FIG. 16 is a schematic diagram illustrating conversion of a pixel value.

FIG. 17 is a schematic diagram illustrating an interleaving process of the divided ALU system.

FIG. 18 is a schematic diagram illustrating conversion into 24-bit color.

[Explanation of symbols]

 REFERENCE SIGNS LIST 100 polygon data input device 200 data processing device 201 polygon data conversion device 202 preprocessing device 203 pixel value interpolation device 204 hidden surface elimination device 205 pixel drawing device 206 pixel value conversion device 300 storage device 301 intermediate information memory 302 intermediate drawing memory 303 frame Buffer 400 Display device 500 Data processing device 501 Preprocessing device 502 Pixel value interpolation device 503 Hidden surface removal device 504 Pixel value conversion / drawing device 600 Storage device 601 Z buffer 602 Frame buffer

Claims (4)

[Claims] 1. A three-dimensional graphics drawing apparatus for displaying an image of a three-dimensional object on a two-dimensional screen using a parallel processing of a divided ALU system, wherein one or a plurality of polygons are drawn for each scanning line. Pixel value interpolation, hidden surface elimination, and drawing within a scanning line are performed using an intermediate drawing storage unit, and the intermediate drawing storage unit stores the pixel value and the Z value in a format suitable for the parallel processing of the divided ALU system. A three-dimensional graphics drawing apparatus, comprising: a conversion unit that stores and, after drawing one or a plurality of scanning lines, converts a pixel value stored in the intermediate drawing storage unit and writes the converted pixel value into a frame storage unit. 2. A three-dimensional graphics rendering apparatus for displaying an image of a three-dimensional object on a two-dimensional screen by using a parallel processing of a divided ALU system, wherein the input polygon data is rendered for each scanning line. A polygon data conversion means for converting the parameters into the parameters necessary for storing the intermediate information storage means, and using the parameters for polygon drawing stored in the intermediate information storage means, Preprocessing means for obtaining and updating the contents of the intermediate information storage means; and dividing A using the parameters supplied from the preprocessing means.
Pixel value interpolation means and hidden surface erasing means for performing pixel value interpolation and hidden surface elimination for drawing one scanning line using LU parallel processing, and intermediate drawing in a format suitable for divided ALU system parallel processing A pixel drawing means for writing a pixel value and a Z value in the storage means; and a pixel value for one scanning line stored in the intermediate drawing storage means after completion of drawing of one or a plurality of scanning lines. A three-dimensional graphics rendering apparatus, comprising: the pixel value conversion means for writing a pixel value at a scanning line position to be written. 3. A three-dimensional graphics rendering device for displaying an image of a three-dimensional object on a two-dimensional screen by using a parallel processing of a divided ALU system, wherein the input polygon data is rendered for each scanning line. Polygon data converting means for converting parameters required by the polygon data converting means, intermediate information storing means for storing parameters obtained by the polygon data converting means in association with each scanning line on a screen, and stored in the intermediate information storing means. Preprocessing means for obtaining parameters necessary for rendering for each scanning line using parameters for rendering polygons, and updating the contents of the intermediate information storage means; and a pixel value for each pixel of one or more scanning lines. And Z
Intermediate drawing storage means for storing values, pixel value interpolation means for obtaining pixel values of pixels in a region occupied by polygons in a scanning line using the parameters obtained by the preprocessing means, and parameters obtained by the preprocessing means Is used to determine the Z value of the pixel in the area occupied by the polygon in the scanning line, and the Z value stored at the position of the corresponding pixel in the intermediate drawing storage means is obtained.
A hidden surface erasing unit that performs hidden surface erasing processing using a value, and a pixel value obtained by the pixel value interpolating unit and a result of the hidden surface erasing process by the hidden surface erasing unit. A pixel drawing unit for drawing pixel data; a frame storage unit for storing pixel values of all pixels on one screen; and a pixel storage unit for reading pixel values of each pixel of the one or more scanning lines from the intermediate drawing storage unit. Pixel value conversion means for performing pixel value conversion into a format suitable for recording to the corresponding pixel, and writing the pixel value to a corresponding scanning line position of the frame storage means. The hidden surface elimination means performs hidden surface elimination by using parallel processing of a divided ALU method, and the pixel drawing means stores the intermediate drawing memory in a format suitable for the parallel processing of a divided ALU method. hand The pixel value conversion unit reads out the pixel value in a format suitable for the parallel processing of the divided ALU system stored in the intermediate drawing storage unit after the drawing of the one or more scanning lines is completed. A three-dimensional graphics rendering apparatus, which converts the data into a format suitable for recording in the frame storage means and writes the converted data at a corresponding scanning line position. 4. Polygon data conversion means for converting input polygon data into parameters necessary for drawing each scanning line, and associating parameters obtained by said polygon data conversion means with each scanning line on a screen. Using the intermediate information storage means for storing and storing, and the parameters for drawing polygons stored in the intermediate information storage means, obtaining parameters necessary for drawing for each scanning line, and updating the contents of the intermediate information storage means Pre-processing means, and for each pixel of one or more scanning lines, the pixel value and Z
An intermediate drawing information storage unit for storing a value, a pixel value interpolation unit for obtaining a pixel value of a pixel of an area occupied by a polygon in a scanning line by using the parameters obtained by the preprocessing unit, and a pixel value obtained by the preprocessing unit. Using the parameters, the Z value of the pixel in the area occupied by the polygon in the scanning line is obtained, and the hidden surface erasing process is performed using the Z value stored at the position of the corresponding pixel in the intermediate drawing information storage means. An erasing unit; a pixel drawing unit that draws a pixel value in the intermediate drawing information storage unit by using a pixel value obtained by the pixel value interpolation unit and a hidden surface erasing process result by the hidden surface erasing unit; A frame buffer storage unit for storing pixel values of all pixels; and a pixel value of each pixel of the one or more scanning lines read from the intermediate drawing information storage unit, and suitable for recording in the frame buffer storage unit. Pixel value conversion means for performing pixel value conversion in a format according to the above, and writing the pixel value to a corresponding scanning line position of the frame buffer storage means, wherein the pixel value interpolation means performs pixel value interpolation using parallel processing of a divided ALU system. The hidden surface erasing means performs hidden surface erasing by using the parallel processing of the divided ALU method, and the pixel drawing means stores the pixel value in the intermediate drawing information storage means in a format suitable for the parallel processing of the divided ALU method. Writing, the pixel value conversion means reads out pixel values in a format suitable for parallel processing of the divided ALU system stored in the intermediate drawing information storage means after drawing of the one or more scanning lines is completed, and In the three-dimensional graphics drawing apparatus, the polygon data conversion unit converts the data into a format suitable for recording in a buffer storage unit and writes the converted data into a corresponding scan line position. Each of the pre-processing unit, the pixel value interpolating unit, the pixel drawing unit, and the pixel value converting unit can be read by a computer recording a program that causes the computer to function on the computer constituting the three-dimensional graphics drawing device. recoding media.