KR100529308B1 - Method and apparatus for detecting film mode in image signal processing - Google Patents

Abstract

The present invention relates to a signal processing method and apparatus, and more particularly, to a film mode detection method and apparatus for image signal processing for determining whether an image signal to be processed is a film mode image converted from a film image to a video image. will be.

According to the present invention, by subdividing a screen into several regions, film mode determination is performed for each region, and signal processing is performed using the film mode determination results for each region, whereby different film mode determination results are generated for each region. In this case, the image quality is deteriorated when signal processing is incorrectly determined. In addition, a signal processing suitable for the film mode determination result is performed for each region to obtain an optimum image quality. . In particular, even if there is text with motion in the film mode image, there is an effect that can prevent the deterioration of the image quality.

Description

Method and apparatus for detecting film mode in image signal processing

The present invention relates to a signal processing method and apparatus, and more particularly, to a film mode detection method and apparatus for image signal processing for determining whether an image signal to be processed is a film mode image converted from a film image to a video image. will be.

In general, an image film produced for movie screening has a specification of 24 frames / sec, and a general interlaced television image has a specification of 60 field / sec. Accordingly, in order to display a film image on a television, a standard conversion is required.

As a method of converting a film image into an interlaced television image, a method of generating two or three field images from an image of one frame is used as shown in FIG. 1. Since the image standard conversion is performed after being performed by a broadcasting station, the television receiving the broadcast cannot determine whether the film image is converted into an interlaced television image or a general interlaced image with no standard conversion. There was no.

In addition, the image processing circuit of the television performs progressive scan conversion to increase the horizontal frequency of the interlaced video signal in order to improve image quality of the video signal or frame rate conversion to increase the vertical frequency. Rate Conversion).

However, in the case of progressive scan conversion or frame rate conversion of the film mode video signal converted from the film image to the interlaced television image, the motion in the image which is virtually motionless in general video processing by the interpolation method by the motion detection is performed. When the interpolation between adjacent frames is recognized as the occurrence of the intermittence, the image quality deteriorates due to the characteristics of the film image.

In order to improve this problem, it was necessary to determine whether the video signal received from the television is a film mode video signal or a general video signal.

Accordingly, the developed film mode discrimination method of the related art calculates the sum of the absolute values of the differences between adjacent frames, so that when the sum of the absolute values of the calculated frame differences exceeds the threshold, it is not '1'. In this case, if the motion detection value is determined as '0', when the determined motion detection value is the film mode, it will be expressed as ... 111101111011110 .... on the frame conversion characteristics as shown in FIG. If the sum of the absolute values of the differences between adjacent fields is calculated, and the motion detection value is determined to be '1' if the sum of the absolute values of the calculated field differences exceeds a threshold, otherwise, '0'. When the determined motion detection value is the film mode, it will be expressed as ... 001010010100101 .... Accordingly, according to the prior art, the determination of the film mode was determined according to whether the calculated motion detection values have such periodicity.

However, in the case where a film mode image signal is transmitted in a film mode when a broadcast station carries a moving text image such as a broadcast time guide together with the film mode image, the film mode determination method according to the conventional technology is applied. There is a problem in that it is incorrectly determined that the image quality is deteriorated by judging and interpolating the film mode image as a general video during progressive scan conversion or frame rate conversion.

The technical problem to be solved by the present invention is to determine the film mode for each area by dividing the entire screen into a plurality of areas in order to solve the above problems, in the image signal processing to determine the final film mode by combining the determination results To provide a film mode detection method and apparatus.

In order to achieve the above technical problem, the film mode detection method in the image signal processing according to the present invention includes (a) subdividing one unit screen into a plurality of regions, and calculating motion detection information for each region. (B) calculating a change pattern of the motion detection information for each area between the previous unit screen and the current unit screen, and (c) changing the pattern of change of the motion detection information for each area by comparing with the change pattern unique to the film mode. And determining the film mode for each region.

In order to achieve the above technical problem, the film mode detection apparatus in the image signal processing according to the present invention uses the input image of the current frame and the image delayed by one frame to determine the motion detection determination value in units of pixels. Motion detection means for calculating, motion detection counting means for each area for calculating motion detection information for each area obtained by adding motion detection determination values of the pixel unit to a plurality of areas, and motion detection information for each area between a current frame and a previous frame delayed by one frame Region-specific motion change determining means for generating a change pattern of the film mode and film mode determination means for determining the film mode for each area when the change pattern of the motion detection information for each area coincides with the change pattern inherent to the film mode. It is characterized by.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

As shown in FIG. 2, the film mode detection apparatus in the image signal processing according to the present invention includes a low pass filter 1, 2 (LPF1, 2; 201, 202), a dispersion calculator 1, 2 (203, 204), and a frame. Delay unit 205, motion determining unit 206, average calculating unit 214, motion detection counting unit 207 for each area, motion change determining unit 208 for all areas, area A, B, C motion change determining unit ( 209, 210, and 211, film mode determination unit 212, and standard conversion unit 213.

Low pass filter 1, 2 (LPF1, 2; 201, 202), distributed computing units 1, 2 (203, 204), frame delay unit 205, motion determining unit 206, and averaging unit 214 among the above-mentioned construction means. Block containing the motion detection block 1000 is referred to as a motion detection block 1000, and a block including the full area motion change determining unit 208 and the areas A, B, and C motion change determining units 209, 210, and 211 is moved by area. This is called a change decision block 2000.

According to an embodiment of the present invention, as shown in FIG. 6, the screen of one frame is subdivided into regions A, B, C, and global regions including regions A, B, and C. Detect the mode. The present invention is not limited to the embodiment in which the region is divided into four regions as described above, and it is natural that the invention can be applied to the embodiment divided into two or more regions.

The input video and the delayed video signal of one frame in the frame delay unit 205 are input to the motion determining unit 206 through the low pass filters 1,2 and 201 and 202 to remove noise in the high frequency band, respectively.

In addition, although the threshold value used for the motion determination may be determined as a fixed initial setting value, in the present invention, in order to minimize the influence of noise, as shown in FIG. 3, the dispersion value of the input image and the one-frame delayed image is used. It is decided variably by a flowchart.

First, the variance values var1 and var2 of the video signal currently input by the dispersion calculator 1 203 and the dispersion calculator 2 204 and the video signal delayed by one frame are respectively calculated (step 301).

Then, the average calculating unit 214 calculates an average value var of the variance value var1 of the input video signal of the current frame and the variance value var2 of the video signal delayed by one frame (step 302).

Next, the motion determination unit 206 executes the following threshold determination process.

First, the variance average value var and the first threshold value vth1 (8 as an example) are compared (step 303).

As a result of the comparison determination in step 303, when the variance average value var is smaller than the first threshold value vth1, the first threshold value vth1 is determined as the motion detection threshold value Mth (step 304).

However, if the variance average value var is greater than or equal to the first threshold value vth1, the variance average value var is compared with the second threshold value vth2 (where vth2> vth1, as an example 16) (step 1). 305).

As a result of the comparison judgment in step 305, when the variance average value var is larger than the second threshold value vth2, the second threshold value vth2 is determined as the motion detection threshold value Mth (step 306).

However, when the comparison average result of step 305 determines that the variance average value var is less than or equal to the second threshold value vth2, the variance average value var is determined as the motion detection threshold value Mth (step 307). That is, when the variance average value var exists between the first threshold value vth1 and the second threshold value vth2, the variance average value var is applied as it is as the motion detection threshold value Mth.

Accordingly, the threshold value for motion detection varies according to the average value of the variance of the video signal of the current frame and the video signal delayed by one frame. This is because when the average value of the variance is large, the video signal contains a lot of noise. When the detection threshold is determined to be high and the average value of the variance is small, the video signal includes less noise. Therefore, the detection threshold is set to be low.

Using the motion detection threshold determined as described above, the motion determination unit 206 calculates the absolute value of the difference between the input video of the current frame and the video signal delayed by one frame in units of pixels, and the calculated absolute value is used for motion detection. If it is larger than the threshold value, it is determined that the motion is detected and the motion determination value 1 is generated.

Then, the motion detection counting unit 207 for each area counts the number of pixels for which the motion determination value 1 is generated for each area in the subdivided area as shown in FIG. 6.

Next, the area-specific motion change determination block 2000 determines the area-specific motion change as follows.

First, a change in motion between frames in a global region is determined by the flowchart of FIG. 4.

First, the motion determination value 1 in the entire area of the current frame generated by the motion detection counting unit 207 for each area counts the number of generated pixels G and the entire area of the previous frame 1 frame. The value PG counting the number of pixels for which the motion determination value 1 is generated is calculated, respectively (step 401).

Next, the whole area motion change determination unit 208 executes the following process.

That is, the G and PG values obtained by adding the number of pixels having a motion determination value of 1 in the current frame and the previous frame are compared with the motion determination threshold value Gth in units of frames (step 402).

As a result of the determination in step 402, when both the G and PG values are smaller than Gth, the output Mi of the full-range motion change determining unit 208 is determined to be 1 (step 403).

If at least one of the G and PG values is greater than Gth as a result of the determination in step 402, the value of G obtained by multiplying PG by a constant coefficient α smaller than 1 is compared (step 404). The reason for comparing with G after multiplying the coefficient α is to accurately detect the change from the image of the previous frame that is not affected by noise to the image of the current frame which is not in motion.

As a result of the determination in step 404, the output Mi is determined to be 0 when G is less than or equal to the value of α × PG (step 405), and otherwise determined to 1 (step 403).

Accordingly, the output Mi of the entire region motion change determining unit 208 becomes 0 only when the image of the previous frame moves from the image having the movement of the previous frame to the image without the movement of the current frame.

Secondly, the motion change between frames in areas A, B, and C is determined according to the flowchart shown in FIG.

First, the value L counting the number of pixels for which the motion determination value 1 is generated in the current frame in each area (areas A, B, and C) generated by the motion detection counting unit 207 for each area, and one frame before. In operation 501, a value PL of counting the number of pixels for which the motion determination value 1 is generated for each corresponding region of the frame is calculated.

Next, the area A, B, and C motion change determining units 209, 210, and 211 execute the following processes, respectively.

That is, the L and PL values obtained by adding the number of pixels of the motion determination value 1 calculated in the current frame and the previous frame for each region are compared with the motion determination threshold Lth in the unit of the frame for each region (step 502).

As a result of the determination in step 502, when both L and PL values are smaller than Lth, the output Mi of the motion change determination units 209 to 211 for each region is determined to be 1 (step 503).

If at least one of the L and PL values is greater than Lth as a result of the determination in step 502, the value of L multiplied by PL multiplied by the variable coefficient beta i smaller than 1 (step 504). The reason for multiplying the variable coefficient beta i and comparing it to L is to accurately detect the change from the image of the previous frame which is not affected by the noise to the image of the current frame without movement. In particular, the variable coefficient beta i varies according to the value of PL. When the PL value is smaller than the predetermined value, the variable coefficient beta i is also decreased. When the PL value is larger than the predetermined value, the variable coefficient beta i is also increased. This is to minimize the influence of the noise because it is easily affected by the noise in the partial region compared to the entire region.

If L is less than or equal to the beta i x PL value, the output Mi of the motion change determination unit 209 to 211 for each region is determined to be 0 (step 505). (Step 503).

Accordingly, the output Mi of the motion change determination units 209 to 211 for each region becomes 0 only when the image changes from the moving image to the motionless image, and 0 otherwise.

Then, by using the region-specific motion change pattern generated in the region-specific motion change determination block 2000 as described above, the film mode determination unit 212 finally has a motion change pattern (.. .1111011110 ....), and in the case of having a motion change pattern inherent to the film mode, the area is determined to be the film mode.

The film mode determination unit 212 generates a single determination result when all regions and all partial regions A, B, and C are determined to be the film mode or the normal mode in the same manner. If different determination results are generated for each region, the film mode determination unit 212 may program to determine all regions as the film mode as an example.

However, as another embodiment, the film mode determination unit 212 outputs different determination results for each area to the standard conversion unit 213, and uses the film mode results determined differently for each area in the standard conversion unit 213 to thereby use the same. It may be designed to perform appropriate signal processing.

That is, as shown in FIG. 7, when the regions A, B, and C are determined to be in the normal mode, and the entire region is determined to be the film mode, the standard conversion unit 213 has a predetermined range around the boundary line where the film determination is different. In the region determined as the normal mode, the buffer region is set in the normal mode, and the image signal standard by the normal video average interpolation processing is converted. In the region determined as the film mode, the image signal is changed by simple insertion of the previous field image rather than the average interpolation. Convert the specification. In the buffer region, the standard conversion is performed by a method in which the weight of the film mode signal processing is higher as the proximity to the film mode region is given, and the weight of the normal mode signal processing is higher as the proximity to the film mode region is obtained.

Accordingly, the film mode detection area and the application area are separated into several areas in the screen, so that even if there is moving text in the film mode image, not the film mode, the signal mode is optimally processed using the film mode determination value for each area. Now you can run

The invention can be practiced as a method, apparatus, system, or the like. When implemented in software, the constituent means of the present invention are code segments that necessarily perform the necessary work. The program or code segments may be stored in a processor readable medium or transmitted by a computer data signal coupled with a carrier on a transmission medium or network. Processor readable media includes any medium that can store or transmit information. Examples of processor-readable media include electronic circuits, semiconductor memory devices, ROMs, flash memories, E ² PROMs, floppy disks, optical disks, hard disks, optical fiber media, radio frequency (RF) networks, and the like. Computer data signals include any signal that can propagate over transmission media such as electronic network channels, optical fibers, air, electromagnetic fields, RF networks, and the like.

Specific embodiments shown and described in the accompanying drawings are only to be understood as an example of the present invention, not to limit the scope of the invention, but also within the scope of the technical spirit described in the present invention in the technical field to which the present invention belongs As various other changes may occur, it is obvious that the invention is not limited to the specific constructions and arrangements shown or described.

As described above, according to the present invention, a film mode determination result is generated by subdividing one screen into several regions to perform film mode determination for each region and signal processing using the film mode determination result for each region, thereby generating different film mode determination results for each region. In this case, the film mode is incorrectly judged as a normal mode, thereby improving the problem of deterioration in image quality during signal processing, and the optimum image quality can be obtained by performing a signal processing suitable for the film mode determination result for each region. That effect is produced. In particular, even if there is text with motion in the film mode image, there is an effect that can prevent the deterioration of the image quality.

1 is a diagram illustrating a method of converting a film image into a video image.

2 is a block diagram of a film mode detection apparatus in the video signal processing according to the present invention.

3 is a flowchart of a method for determining a motion detection threshold used in determining a motion according to the present invention.

FIG. 4 is a flowchart of a method for determining a change in full region motion by the configuration diagram shown in FIG.

FIG. 5 is a flowchart of a method for determining a change in motion for each partial region according to the configuration shown in FIG. 2.

6 illustrates an example of splitting a screen area according to an embodiment of the present invention.

7 is a diagram for explaining signal processing in the case where the film mode determination is different for each region.

Claims (19)

In the video signal processing method, (a) a threshold value for motion detection, in which a current unit screen and a previous unit screen are divided into a plurality of regions, and the difference between signals of the current unit screen and the previous unit screen for each of the divided regions is variable according to noise of the unit screen; Calculating a motion detection determination value in units of pixels in comparison with; (b) calculating motion detection information for each area by adding the motion detection determination values for each area; calculating a change pattern of the motion detection information for each area; And and (d) if the change pattern of the motion detection information for each region coincides with the change pattern inherent to the film mode, determining the corresponding region as the film mode. . The method of claim 1, wherein the unit screen is a frame unit screen.  The method of claim 1, wherein the unit screen is a field unit screen. delete The method of claim 1, wherein the motion detection threshold is varied according to an average value of a dispersion value of an image of a current unit screen and a dispersion value of an image of a previous unit screen. 6. The method of claim 5, wherein the threshold value is determined as a threshold value when the average value of the dispersion values is smaller than a predetermined first threshold value, and the predetermined second threshold value (greater than the first threshold value). Value), the second threshold value is determined as the threshold value, and when present between the first threshold value and the second threshold value, the average value of the dispersion values is determined as the threshold value. Film mode detection method in. The method of claim 1, wherein the change pattern of the motion detection information for each area is compared with motion detection information for each area between a previous unit screen and a current unit screen in a plurality of consecutive unit screens. And pattern information obtained by generating a first logical value under changing conditions and a second logical value (value inverted by the first logical value) under other conditions. The method of claim 1, wherein the change pattern of the motion detection information for each region comprises at least one of motion detection information of a previous unit screen and a current unit screen larger than a predetermined threshold, and the motion detection information of the current unit screen is increased. The first logical value is generated only in a condition smaller than the product of motion detection information of the previous unit screen multiplied by a predetermined coefficient value of 1 or less, and in other conditions, the second logical value (the value whose first logic value is inverted) is generated. Film mode detection method in the image signal processing, characterized in that the obtained pattern information. 10. The method of claim 8, wherein the count value is varied according to the size of motion detection information for each region. The method of claim 1, wherein the determination of the film mode determines all regions as the film mode when it is determined that the film mode is in at least one or more regions. The method according to claim 1, wherein when the determination of the film mode differs for each of the regions, a predetermined buffer region is set at the boundary of each region, and in the buffer region, the application weights of the normal mode and the film mode are different depending on the distance from the region. And a progressive scan conversion or frame rate conversion process applied to the film mode detection method. In the video signal processing apparatus, The input image of the current frame and the image delayed by one frame are respectively divided into a plurality of areas, and the difference in signal between the input image of the divided current frame and the image delayed by one frame is variable according to the noise of the image signal. Motion detection means for generating a motion detection determination value on a pixel basis in comparison with; Area-specific motion detection counting means for calculating area-specific motion detection information obtained by adding the motion detection determination value in units of pixels to each divided area; Area-specific motion change determining means for calculating a change pattern of the area-specific motion detection information; And And film mode determination means for determining the corresponding region as the film mode when the change pattern of the motion detection information for each region coincides with the change pattern inherent to the film mode. The method of claim 12, wherein the motion detecting means A low pass filter 1 for blocking high frequency noise of an input video signal; A frame delay unit for delaying the input video signal by one frame; A low pass filter 2 for blocking high frequency noise of the one-frame delayed video signal output from the frame delay unit; A dispersion calculator 1 for calculating a dispersion value of an input video signal; A dispersion calculator 2 for calculating a dispersion value of the video signal delayed by one frame; An average calculating unit for calculating an average value of the dispersion values generated by the dispersion calculating units 1 and 2; And A threshold value corresponding to a magnitude of an average value calculated by the average calculating unit is generated, and an absolute value of a pixel unit difference between an image signal output from the low pass filter 1 and an image signal output from the low pass filter 2 is determined by the threshold value. And a motion determination unit for generating a motion detection determination value when the value is larger than the value. The method of claim 13, wherein the threshold value is determined as a threshold value when the average value of the variance values is smaller than a predetermined first threshold value, and the predetermined second threshold value (greater than the first threshold value). Value), the second threshold value is determined as the threshold value, and when present between the first threshold value and the second threshold value, the average value of the dispersion values is determined as the threshold value. Film mode detection device in the device. 15. The method of claim 14, wherein the change pattern of the motion detection information for each area is compared with motion detection information for each area between a previous frame screen and a current frame screen in a plurality of consecutive frame screens, and the motion-free image is a motionless image. And pattern information obtained by generating a first logical value under changing conditions and a second logical value (value inverted by the first logical value) under other conditions. The method of claim 13, wherein the change pattern of the motion detection information for each region comprises at least one of motion detection information of a previous unit screen and a current unit screen being greater than a predetermined threshold, and the motion detection information of the current unit screen is increased. The first logical value is generated only in a condition smaller than the product of motion detection information of the previous unit screen multiplied by a predetermined coefficient value of 1 or less, and in other conditions, the second logical value (the value whose first logic value is inverted) is generated. Film mode detection apparatus in the image signal processing, characterized in that the acquired pattern information. 17. The film mode detection apparatus of claim 16, wherein the count value is varied according to the size of motion detection information for each region. The film mode detection apparatus according to claim 13, wherein the film mode determination means determines all regions as the film mode when it is determined that the film mode is in at least one or more regions. The method according to claim 13, wherein when the determination of the film mode differs for each of the regions as a result of the determination by the film mode determining means, a predetermined buffer region is set at the boundary of each region, and in the buffer region, application weights of the normal mode and the film mode are applied. And a standard conversion unit for performing progressive scan conversion or frame rate conversion processing by applying differently depending on the distance from the area.