KR20030073653A - Conversion method for scanning mode of television - Google Patents

Abstract

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of converting a scanning method of a television. In the related art, it is not possible to know whether a source of an encoding stream is a film source or a general video source, so that the display is displayed by interlaced scanning or progressive scanning regardless of the type of the source. In the process of displaying in a progressive scan method using a film source of a progressive scan method, there is a problem that the image quality is reduced. In view of the above problems, the present invention includes the steps of: converting an encoded video stream of 24 frames into a frame of a method of scanning 30 frames per second by 3: 2 pulling down each frame; Detecting a time difference between fields of the same position of adjacent frames in the frame, determining the regularity of the time difference, and determining whether the source of the encoded video stream is film or interlaced video; Performing an appropriate weave according to the source type of the determined encoded video stream; And displaying the weaved frame in a scanning manner corresponding to the source, and determining whether the source of the encoded video stream is film or interlaced video, and performing appropriate de-interlacing for each source, When displaying a video stream sourced with a film on the display device, there is an effect of preventing deterioration of the picture quality.

Description

Scanning method conversion method of TV {CONVERSION METHOD FOR SCANNING MODE OF TELEVISION}

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of converting a scanning method of a television, and more particularly, to a method of converting a scanning method of a television, which detects a format of an interlaced scanning input image and converts it into a sequential scanning method according to the detection result. .

In general, the scanning method of television is divided into interlaced and progressive modes.

Sequential scanning used in computer monitors and digital televisions simultaneously displays all horizontal scan lines in one frame, while interlaced scans used in standard TV formats NTSC, PAL and SECAM display only half of the horizontal scan lines at a time.

In other words, the interlaced scan is displayed after the first field of odd number is displayed and the second field, which is an even number of horizontal scanning lines, is used to take advantage of the persistence of vision (afterimage) characteristic of the human eye.

This interlaced feature is that half of the data can be used to achieve high refresh rates (50 or 60 Hz).

However, there is a disadvantage that the resolution is inevitably reduced by half because the video must pass through the filter to prevent factors such as shaking.

The difference between the interlaced scan and the progressive scan is determined during the imaging process of the image source. That is, the film camera shoots 24 frames per second, and the video camera alternately captures odd and even fields at 1/60 second intervals.

As a result, the film mode is used to sequentially scan 24 frames per second to display a movie.

Many movies are now available on DVD discs. The current DVD standard was established assuming that 30 frames per second were scanned by interlaced scanning.

However, in the case of a movie, a method of scanning 24 frames per second, which is a source of a sequential scanning method, due to the difference of the imaging method, is required to convert a frame rate when encoding.

The conventional method of converting the frame rate in encoding is to encode 24 frames while interpolating 30 frames per second on the display using the refit first field and top field first field in the MPEG stream standard. have.

In addition, as another method, 24 frames are pulled down to 30 frames 3: 2, and then encoded and used. Hereinafter, such conventional methods will be described in detail with reference to the accompanying drawings.

1A to 1C are each applied to an encoding stream of 24 conventional frames and an interlaced display device that decodes the stream of 24 frames and scans 60 fields per second, or scans 60 frames per second of the 24 frames. The result of performing proper decoding on the progressive scanning method is shown.

First, 1, 1, 0, 0, 0, 1, 0 are applied to the top field first to the encoding streams I1, P2, B3 to B7 of 24 frames shown in FIG. 1A, and the refit first fields are respectively 1 and 0. , 0, 1, 1, 1, 0 applies.

The I1, P2, and B3 to B7 each have a frame unit interval (FRAME UNITINTERVAL, FU) set as a transmission order (TRANSMISSION ORDER). This is for synchronizing the audio signal with the video signal. The transmission unit can synchronize with the audio signal by appropriate combination of transmission orders, and the frame unit interval of each transmission order is slightly different according to each standard.

The top field first (TOP_FIELD_FIRST) and the refit first field (REPEAT_FIRST_FIELD) applied to the encoding stream are defined in the MPEG grammar, and the interlaced display device or the sequential scanning method that scans 60 fields per second and 60 frames per second using the same. The signal is converted into a signal suitable for the display device.

FIG. 1B is a diagram illustrating decoding of the 24 frame encoded streams for use in an interlaced display device that scans 60 fields per second, and tops the frames of the 24 frame encoded streams I1, P2, B3 to B7. It divides into bottom (TOP, BOTtom) fields so that 30 frames per second can be displayed by interlaced scanning.

In this state, an encoding stream of 24 frames per second is applied, but the actual display process can scan and display 30 frames per second by interlaced scanning.

FIG. 1C is a decoding result for applying the encoded stream of 24 frames to a sequential scan type display device scanning 60 frames per second. As shown in FIG. 1C, the transmission order may be changed to synchronize with an audio signal. Will be.

Since the above method has a FLAG that each frame is derived from a sequential scanning method source having 24 frames in the MPEG stream grammar, it can be easily displayed as a sequential scanning method scanning 60 frames per second.

2A to 2C are other exemplary embodiments of the present invention. As shown in FIG. 2A to 2C, 24 frames F1 to F4 captured in the film mode are applied to each other, and each frame is divided into a top and a bottom to use 30 frames per second. Convert to a sequential scan stream.

This encodes the transmission orders I1, P2, B3 to B7 as shown in Fig. 2C, and applies all of the top field first to 1 and all of the refit first fields to zero.

In the display device which interlaces 60 fields per second of the encoding stream, the fields of each frame may be sequentially displayed.

Also, in the method of sequentially scanning 60 frames per second using the encoding stream as shown in FIG. 2C, there is no time difference information for each field in the frame in the encoding stream, and thus there is no difference from the stream that interlaces 30 frames per second. There is no.

In the case of scanning such a stream in a sequential scanning method, it is not known that it is a film source at the time of image capturing, and is displayed by normal de-interlacing, so that the image quality is deteriorated by the difference of the method.

As described above, in the conventional television scanning method, it is not possible to know whether the source of the encoding stream is a film source or a general video source, and the progressive scanning method is performed by displaying in interlaced scanning or progressive scanning regardless of the type of the source. There was a problem that the image quality is degraded in the process of displaying in a sequential scanning method using a film source of the method.

In view of the above problems, the present invention provides a method of converting a scanning method of a television which can prevent the deterioration of image quality by determining the type of the encoding stream and performing appropriate decoding thereof when displaying a film source of a sequential scanning method. There is a purpose.

1A to 1C are schematic diagrams of a process of converting an encoding stream of a conventional 24 frame into a format suitable for interlaced scanning or sequential scanning.

Figure 2a to 2c is another embodiment of a conventional scanning method conversion method.

Figures 3a to 3c is a schematic diagram showing a method of de-interlacing (DE-INTERLACING) of the interlaced source stream made from the film applied to the present invention.

4 is a schematic diagram showing the regularity of an encoding stream of a film source.

5 is a conceptual diagram of de-interlacing.

6 is a flowchart of a process of detecting a source of an encoding stream in the present invention.

Figure 7 is a schematic diagram showing the weave process of the film source applied to the present invention.

8 is a schematic diagram showing a process of the present invention when a film source and an interlace source are continuously applied to a video stream.

The above object is to determine the source of the encoding stream by using the regularity of the time difference between fields in each adjacent frame of the encoding stream obtained from the film source, and to decode accordingly according to the type of the determined source. This is achieved by performing the present invention described in detail with reference to the accompanying drawings as follows.

3A to 3C are schematic diagrams showing the present invention's DE-INTERLACING method of interlaced source streams made from film, as shown in FIG. 3A to FIG. Separate the top and bottom fields (TOP, BOT) from (F1 ~ F4) into a stream of 30 frames per second (Figure 3b), and reconstruct the topfield and bottomfield by finding the original frame fields. In other words, it converts a stream of 60 frames per second (FIG. 3C).

In the above process, when each frame is divided into a top field and a bottom field, it is necessary to determine whether the source of the encoded stream from which the divided state is derived is a film source or a video source. It is important to obtain a scanning stream corresponding to the weave of the top field and the bottom field at the correct position.

FIG. 4 is a schematic diagram of a method of detecting a film mode when the encoding source is a film. As shown in FIG. 4, an adjacent frame is applied when an encoding stream of 24 frames / sec is applied and 3: 2 pulls down to make 30 frames. Difference values of the same field of have constant regularity.

That is, the time difference between the top field of the first frame F0 and the top field of the next frame F1 is small, and the difference of the bottom field between the frames is large and the value of the frame F1 and the next frame F2 is large. The difference between the top field and the bottom field is large, and the comparison results through this process have a specific pattern.

That is, when the time difference is relatively small, S, and the time difference is relatively large, L, 'SLLLL' appears repeatedly.

Such a pattern has certain regularity when in film mode, and the present invention uses this regularity to determine whether the encoding source is film.

The concept of de-interlacing for the determination of such regularity is described first.

FIG. 5 is a conceptual diagram of basic de-interlacing. As shown in FIG. 5, due to the time difference between fields included in the frame D_FRAME, the current field and the next field should not be simply combined when creating one frame.

At this time, the application is performed by using motion estimation or edge detection.

The CUR_FIELD is a previously deinterlaced frame, a future field, or current field data. The CUR_FIELD is used to predict the next field and restore an empty field.

However, when the source is a film source, it is better to find the original field where the field is located and perform a WEAVE.

6 is a flowchart of a film mode detection method using the time difference between fields at the same position of the adjacent frame. As shown therein, the minimum value min and the maximum value Max are initially set equal to each other, and the previous state. Set the minimum value of Pseudo_min to the maximum value.

The count values cnt and mcnt are all set to zero.

The count value cnt is a value for counting a large value L of a difference between values located in the same field of the adjacent frame, and the count value m_cnt is a repetition pattern of a small value S and a large value L, That is, it is a value which counts the repetition number of S, L, L, L, and L patterns which show that it is a film source.

Then, a difference (calc_diff) of fields at the same position of the adjacent frame is calculated (calc_diff) and an average (cal_diff_average) of the values is calculated.

Then, if the value obtained by multiplying the set value T1 by the average value diff_average is smaller than the difference value, that is, when the difference value is very large, it is determined that the source of the encoding stream is interlaced video. .

Then, if the difference value (diff) of the time is not very large, it is determined whether the difference between the field value of the first frame and the second frame, that is, the difference value (diff) is smaller than the minimum value and the minimum value (min) Set the difference value (diff) to the initial set value, return to calculate the difference between the second frame and the third frame, and find the average value.

The difference between the field values at the same position among the frames thus obtained is large through the large and small judgment with the minimum value, that is, the difference between the remaining field values based on the difference between the top field values between the first frame and the next frame. Is greater, count the value.

By repeating this process, it is determined whether the counted value is 4 if the remaining difference (diff) is larger than the difference between the top field of the first frame and the second frame, and if it is not 4, the source of the stream is interlaced video. If the value is again at 4, the comparison result is smaller than the previous state, that is, if the determination result follows the above-mentioned pattern of the SLLLL, the count value m_cnt is increased by one.

That is, if each frame is detected from the film, the field value of the first frame and the second frame is small (S), and the field value comparison result up to the fifth frame thereafter is large (L), which is SLLLL. Will follow the pattern of.

As a result of comparing the difference between the field value of the next frame and the minimum value min to determine the regularity, the count value cnt is increased when the difference value diff is greater than the minimum value min. In this case, when the difference value (diff) is smaller than the minimum value (min), since the small value continues twice, it is determined that the frame is not the film mode.

Then, it is determined whether the count value cnt is 4, and when the count value is 4, a difference value is calculated, and the magnitude of the minimum value (Pseudo_min) between the difference value (diff) and the large value (L) of the previous state is determined. If the minimum value (Pseudo_min) among the large value (L) of the previous state is larger, the minimum value is set as a difference value, the count value (cnt) is initialized to 0, and the number of repeated iterations is judged to determine the repeated number of count values (m_cnt). Increase

Through this process, it is determined whether the difference between the same field values between two adjacent frames is small (S) and the large value (L) meets the rule of repeating the SLLLL indicating the film mode, and counts the number of repetitions.

When the number of times T2 of the SLLLL is set, the film mode is enabled, so that the decoding of the film mode can be performed in the decoding process.

The flowchart is to detect a unique rule of the film frame to determine whether the frame is detected in the film, thereby facilitating the determination of the decoding method when performing the decoding.

Summarizing the film mode detection method, the field value of each frame is detected in the initial initialization state, the difference value of the field value is obtained, and the difference value is compared with the minimum value among the larger values (L) among the difference values in the previous state. To determine the magnitude.

At this time, if the current difference value is small, it is designated as S, and if it is high, it is determined whether the values have the rules of SLLLL, and it is judged how many times the SLLLL is repeated. The frame is determined to be a film mode.

After the detection in the film mode as described above, the data of each frame F0 to F4 is reconstructed.

This is to make the field data scanning 60 fields per second into frame data representing 60 frames per second. As shown in Fig. 7, each frame is the current field value of the previous state when configuring the first frame F0. And field of each frame F0 to F4 are combined using the previous field value or the field value of the next state.

In FIG. 7, f0 to f4 refer to a field, and t and b described after the fields f0 to f4 represent a top field and a bottom field.

Most of the weave process uses the current field value and the subsequent field, but constructs the frame data using the previous field as shown by the circle.

Through this process, the film frame that scans 24 frames per second is changed to a sequential scanning method that scans 60 frames per second, and the pattern is detected with regularity in the encoding stream, and the 24 frames are detected. It is possible to detect that it was a film and to perform appropriate decoding on it.

In addition, FIG. 8 illustrates a process in which the type of encoding stream applied while detecting and displaying the film mode is applied by mixing interlaced video encoding.

As shown in Fig. 8, a frame F2 is a mixture of a frame F0 and F1 of an original film, a frame of an original film and a field of an interlaced video, and a frame F3 of an original interlaced video. F4) can be applied continuously.

This means that the video stream is edited by 3: 2 pull-down encoding of the film and the interlaced video encoding. If the film mode is maintained at this boundary, interlaced fields having a large time difference between each field form one frame. While the screen is being combined, the phenomenon occurs.

In order to prevent such a problem, if the difference of each field suddenly increases, it is determined that the original of the encoding stream is not film, and the weave method is changed.

That is, the difference between the top field of the first frame F0 and the second frame F1 is a small value S, and the difference between the bottom field of the first frame F0 and the second frame F1 is a large value L and the second frame ( The difference between the top field of F1) and the third frame F2 is a large value (L), and the difference (diff) of the bottom field value between the second frame (F1) and the third frame (F2) is a very large value (Big_L). In this case, the difference between the bottom field of the second frame F1 and the third frame F2 is greater than the value T1 multiplied by the difference value difference (diff_average) of each field as shown in the flowchart of FIG. 6. Since it becomes large, it is determined as a very large value Big_L.

The difference between these values is an area in which the scene is changed by editing, and the time difference of the scene is large, in particular, the time interval becomes larger by the scanning method, and when the time difference is detected and the value is large, the film When weaving is performed as in the mode, the field having a large time difference is mixed in the same frame and its display characteristics are deteriorated. When the very large value Big_L is detected, the film mode is terminated and interlaced. (weave).

At this time, weave is performed in the film mode on the frame F2, which causes data loss of one field.

However, the human eye cannot find out that one field of data has been lost, and thus deterioration of the image quality cannot be found, which does not cause a problem.

As described above, the present invention obtains a time difference between fields of the same position included in each adjacent frame in an encoded video stream, detects a periodic change pattern of the time difference value, and determines whether the source of the encoded video stream is film or interlaced video. By judging and performing appropriate de-interlacing for each source, there is an effect of preventing deterioration of image quality when displaying a video stream using a film as a source on a sequential scan type display device.

Claims (4)

Converting the encoded video stream of 24 frames into frames of 3: 2 pull-down for each frame and scanning 30 frames per second; Detecting a time difference between fields of the same position of adjacent frames in the frame, determining the regularity of the time difference, and determining whether the source of the encoded video stream is film or interlaced video; Performing an appropriate weave according to the source type of the determined encoded video stream; And displaying the weaved frame in a scanning manner corresponding to the source thereof. Obtaining a time difference between the top field and the bottom field of the adjacent frame; Comparing the time difference with a reference value and determining that the value is less than or greater than the reference value; Counting the number of small cases and large cases to determine the number of occurrences of the small case and the large case, and determining whether there is a pattern in which the case where the number of the arrangements is small and the case where the number of arrangements is large is continuously generated; It is determined whether the pattern occurs continuously more than the set number of times, and when the pattern occurs continuously, it is determined that the source of the encoded video stream is film. When the pattern is not continuous or when the pattern does not occur, the encoded video is determined. Determining the source of the stream as interlaced video; Weaving according to the type of the determined source, and scanning in a suitable scanning method, characterized in that the scanning method of the television. 3. The method of claim 1 or 2, wherein performing the weave when the encoded video stream is a film finds the original position of each field in film mode and constructs a frame with a combination of the fields, the fields of the frame. Is a scanning method of a television, characterized in that combining the field of the current frame and the field before or after. 3. The method of claim 2, further comprising the step of determining the source of the encoded video stream as interlaced video when the time difference between field values of the same position of the adjacent frames is equal to or greater than a set time difference. How to convert.