KR20030008685A - Method of predicting space for preventing error from propagation in image compression system - Google Patents

Abstract

PURPOSE: A method for predicting a space in an image compression system is provided to prevent successive deterioration of quality due to the increase in error. CONSTITUTION: A method for predicting a space to prevent the increase in errors in an image compression system includes the steps of examining a compression mode of a peripheral image block of an image block to be intra-coded within a sheet of image, and confirming propagation of an error of the peripheral image block if the peripheral image block is in an inter mode(301-302); intra-coding the image block by spatially predicting from a peripheral intra image block or a peripheral inter image block to which the error has not propagated(305); and interrupting the spatial prediction from the peripheral inter image block to which the error has propagated(303-304).

Description

Method of predicting space for preventing error from propagation in image compression system}

The present invention relates to a spatial prediction method for preventing error propagation in an image compression system, and to a computer-readable recording medium recording a program for realizing the method. By preventing the prediction encoding from the generated video, it is possible to prevent the propagation of an error.

First, in order to help the understanding of the related art and the technical gist of the present invention, a schematic process of image compression will be described based on the recommendation for image H.261.

There are two screen compression methods, an intra screen compression method and an inter screen compression method.

For intra-screen compression, first, Discrete Cosine Transform (DCT) is performed in 8 × 8 pixel units. This conversion has the effect of concentrating 8 × 8 blocks of energy on low frequency components.

However, since the human visual characteristic feels relatively less quantization noise of high frequency components, the quantization interval is increased as the coefficient of DCT conversion is high frequency.

Therefore, the high frequency component is small in size but large in quantization interval, so that most of them are zero. The quantized DCT coefficients have a large number of zeros. In order to efficiently compress the DCT coefficients, run-length coding and zigzag scan are used. In other words, if 0 is continuous while recording data in the zigzag direction, it is not just to record 0 but only how many times 0 is continuous. If it is not 0, it just records the data. This makes it possible to represent many consecutive zeros with fewer bits at once, resulting in the effect of compressing the data.

Screens created using this method are called "intra-frames" in H.261.

Meanwhile, the inter-screen compression method divides the screen into macroblocks (four luminance blocks and two color blocks) to a certain size unit to increase the compression rate by using the correlation between the screens, and moves the screens at each position of the previous screen. It uses a method called motion compensation to obtain a motion vector indicating whether it is coming and a differential pulse code modulation (DPCM) that encodes and transmits only a difference in an entire image.

Here, the motion compensation method is a motion estimation method. A block matching algorithm is widely used to find a position where the minimum macro error is obtained for each pixel position of a previous screen.

The motion vector is transmitted to the receiving side for screen reconstruction so that the receiving side determines which position in the previous image is the data and decodes it. If the data cannot be solved by motion compensation, the DPCM, which transmits only the difference from the previous image, creates the image that was originally intended to be transmitted.

The screen created by the inter-screen compression method is called "inter-frame".

As such, H.261 consists of Intra-Frame and Inter-Frame, both of which are used together. That is, first, an intra-frame is transmitted to a receiving side, and a much smaller inter-frame is transmitted. This is because it is effective to send an intra-frame when the scene is changed or the movement is large in the transmitted image, and when the motion is small, it is more effective to use an inter-frame.

Currently, in a system for transmitting and receiving video and sound, a video signal and audio signal are encoded and transmitted as a digital signal, stored in a storage unit, and decoded and reproduced. Methods used for encoding such video signals include a transform coding method, a differential pulse code modulation (DPCM) method, a vector quantization method, and a variable length coding (VLC) method. These coding schemes are used to compress the total amount of data by removing redundancy data included in the digital video signal. In order to perform such an encoding method, a screen is divided into blocks having a predetermined size, and a predetermined transformation is performed on each block or a difference signal between blocks to convert image data into a transform coefficient of a frequency domain.

Examples of data transformation methods for each block include DCT (Discrete Cosine Transform), WHT (Walsh-Hadamard Transform), DFT (Discrete Fourier TRansform), and DST (Discrete Sine Transform). The transmission data is compressed by appropriately encoding such transform coefficients according to data characteristics. In order to restore the encoded and transmitted video data to the original video signal, it is necessary to transmit the head data including the encoded state information of the video signal together with the transmission of the video data.

In general, in the variable-length coding scheme of a moving picture, an intra mode is used when there is a large change between images, and an inter mode is used when there is a small change between images. Information about the data is also included in the overhead data.

1 is an exemplary configuration diagram of a general video compression system, and illustrates an example of an inter / intra mode encoding apparatus.

In FIG. 1, image data VS is input to an N × N converter 11 through a first adder A1, and an output terminal of the N × N converter 11 is connected to an input terminal of the quantization unit 12. The output terminal of the quantizer 12 is connected to the variable length encoder 13. On the other hand, the image data VS is also supplied to the mode determination unit 20, so that the mode determination unit 20 determines the inter mode or the intra mode according to the degree of change of the image, and then stores the inter mode / intra mode information in the next bit. Supply to the map converter 21 and the two switches (RSW1, RSW2).

The output terminal of the bitmap converter 21 is connected to a run-length encoder 22, and the output of the run-length encoder 22 is supplied to the variable length encoder 23 so that the variable length coded mode information is provided. Is applied to the buffer 14. In this case, the variable length coded signal is transmitted after being delayed in the buffer 14, and a signal for determining the quantization level QS according to the buffer 14 state is supplied to the quantization unit 12 and the inverse quantization unit 15. do.

Meanwhile, the quantization unit 12 outputs a quantized signal to the inverse quantization unit 15, and the inverse quantization unit 15 is connected to an input terminal of the N × N inverse transform unit 16. At this time, the signal output from the N × N inverse transform unit 16 is applied to the frame memory 17 through the second adder A2, and the frame memory 17 is the copper estimator 18 and the copper compensator 19. Is configured to apply each output. The output of the copper compensator 19 is applied to the first adder A1 and the second adder A2, and the copper estimator 18 receives the image data VS to the copper compensator 19. It is configured to output the vector MV and transmit it together with the encoded video signal VC.

Referring briefly to the operation of the inter / intra mode encoding apparatus having the above configuration, the image data VS of the N × N block (generally, N1 × N2 block) includes the mode determiner 20 and the first adder A1. ) And the same estimator 18. At this time, the mode determination unit 20 encodes each mode in an inter mode when there are many similar parts between the previous picture and the current picture, and in an intra mode when there is a large change between the previous picture and the current picture. Determine how. The inter mode / intra mode determined in this manner determines the inter mode / intra mode of the encoding apparatus by controlling the two switches RSW1 and RSW2.

However, H.26L, which is currently being standardized for the next generation video compression algorithm, uses a spatial prediction method to improve the compression efficiency. In this method, prediction and differential coding are performed based on the image of the surrounding space in a single image. In this case, when an error occurs in one block, the error is propagated to the predicted block based on the error, causing deterioration of image quality.

If an error occurs in the image transmission, an intra encoding method is used in units of macroblocks to prevent propagation of the error. However, when a conventional spatial prediction method is used, an error is detected in neighboring blocks. In the case of propagation, an error propagates up to an intra coded block, thereby reducing the effect of intra coding.

That is, in the case of image compression such as H.26L, if the conventional spatial prediction method is used, there is a problem that continuous deterioration of image quality and image quality deterioration due to propagation of an error are accumulated through prediction encoding from an error image.

The present invention has been proposed to solve the above-mentioned problems, and prevents the error from being propagated by preventing predictive encoding from an image that has an error during image compression such as H.26L. It is an object of the present invention to provide a spatial prediction method for preventing accumulation of deterioration and deterioration of image quality, and a computer-readable recording medium having recorded thereon a program for realizing the method.

1 is an exemplary configuration diagram of a general video compression system.

2A and 2B are explanatory diagrams showing a conventional spatial prediction process.

3 is a flowchart illustrating an embodiment of a spatial prediction method for preventing error propagation according to the present invention.

* Explanation of symbols for the main parts of the drawings

20: mode determination unit 21: bitmap conversion unit

22 run length encoder 23 variable length encoder (VLC)

In order to achieve the above object, the present invention provides a spatial prediction method for encoding an image in an image compression system, wherein the compression mode of the neighboring image block of the image block to be intra-coded in a single image is examined, A first step of checking whether an adjacent image block propagates in an inter mode; A second step of spatially predicting the image block from a neighboring intra image block or a neighboring inter image block not propagated by error; And a third step of blocking spatial prediction from the error propagated peripheral inter image block with respect to the image block.

In addition, the present invention is to check the compression mode of the neighboring video block of the video block to be intra-coded in a single image in the image compression system having a processor, and the error of the neighboring video block when the neighboring video block is the inter mode A first function of confirming propagation; A second function of spatially predicting the image block from a neighboring intra image block or a neighboring inter image block not propagated by error; And a computer readable recording medium having recorded thereon a program for realizing a third function of blocking spatial prediction from an error propagated peripheral inter image block for the image block.

The present invention relates to an improvement method of a spatial prediction method for preventing the error generated when transmitting a video encoded by H.26L to the next video, and to improve the existing spatial prediction method to predict from an error image. If the encoding can be prevented, the propagation of errors can be prevented. As a result, it is possible to prevent continuous deterioration of image quality and accumulation of image quality deterioration due to error propagation.

The above objects, features and advantages will become more apparent from the following detailed description taken in conjunction with the accompanying drawings. Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

In most video compression systems, when an image is encoded, as shown in FIG. 2A, if an error occurs in the C block of the previous image, the error is propagated to the C block, and the C block is used to recover the original image. There is a problem that can not be restored.

In order to prevent this, an intra coding method of a C block is used. In this case, the propagation of an error may be prevented by not predicting the C block. This is called "Intra Refresh".

However, when intra coding is performed in H.26L, as shown in FIG. 2B, a spatial prediction method is used from an upper A block or a left B block. In this case, when the A or B blocks are inter-coded and an error is propagated in the previous image, an error of the previous image is also propagated in the C block, so that the effect of intra refresh cannot be seen. Only the consumption of bit strings becomes large.

Therefore, if the block coding mode of the A or B block is checked, and the inter block is used to block spatial prediction from the block, error propagation to the intra block C can be prevented. Image quality deterioration due to error propagation of an image can be prevented and image quality deterioration due to error propagation and accumulation can be prevented.

3 is a flowchart illustrating a spatial prediction method for preventing error propagation according to the present invention.

As shown in FIG. 3, in the spatial prediction method for preventing error propagation according to the present invention, first, when an image is compressed in an intra mode (301), an adjacent block is an intra block or an inter block. (302). That is, in FIG. 2B, when the C block is compressed in the intra mode, the block encoding mode of the A or B block, which is the neighboring video block, is checked.

As a result of the check, when the neighboring video block is an inter block (that is, inter coded), the neighboring video block checks whether an error is propagated in the previous video (303), and the video block propagated in the error For block 2, block spatial prediction from the block. On the other hand, if the neighboring video block is not an inter-block in which an error propagates or is an intra block (that is, an intra coded), spatial prediction is performed from the corresponding block (305). That is, in FIG. 2B, when the C block is compressed in the intra mode, when the neighboring block A or B is an Inter Block, the error propagation of the A or B block is checked and an error is propagated. Block the spatial prediction for the block. However, if the A or B block is not an error propagated Inter Block or the A or B block is an Intra Block, spatial prediction is performed from the A or B block.

The method of the present invention as described above may be implemented as a program and stored in a computer-readable recording medium (CD-ROM, RAM, ROM, floppy disk, hard disk, magneto-optical disk, etc.).

The present invention described above is not limited to the above-described embodiments and the accompanying drawings, and various substitutions, modifications, and changes are possible in the art without departing from the technical spirit of the present invention. It will be apparent to those of ordinary knowledge.

As described above, the present invention can improve the existing spatial prediction method to prevent predictive encoding from an errored image, thereby preventing the propagation of an error, and continuously deteriorating image quality and image quality deterioration due to error propagation. There is an effect that can prevent the accumulation of.

Claims (4)

A spatial prediction method in image encoding in an image compression system, A first step of checking a compression mode of a neighboring video block of an image block to be intra-coded in a single image and checking whether or not an adjacent image block is propagated when the neighboring video block is an inter mode; A second step of spatially predicting the image block from a neighboring intra image block or a neighboring inter image block not propagated by error; And A third step of blocking spatial prediction from the error propagated peripheral inter image block with respect to the image block Spatial prediction method for error propagation in the image compression system comprising a. The method of claim 1, The video, Substantially, the spatial prediction method for preventing error propagation in an image compression system characterized in that the H.26L image. The method according to claim 1 or 2, The video block, A spatial prediction method for preventing error propagation in an image compression system, which is substantially an intra macro block in an image compressed in an inter mode. In an image compression system having a processor, A first function of checking a compression mode of a neighboring video block of an image block to be intra-coded in a single image and checking whether or not an error is propagated in the neighboring video block when the neighboring video block is an inter mode; A second function of spatially predicting the image block from a neighboring intra image block or a neighboring inter image block not propagated by error; And A third function for blocking spatial prediction from the error propagated peripheral inter image block for the image block A computer-readable recording medium having recorded thereon a program for realizing this.