KR20080035390A - Method and apparatus for deciding video prediction mode - Google Patents

Abstract

A method and an apparatus for determining a video prediction mode are provided to determine a cost value as the standard of prediction mode determination in consideration of SAD and the characteristics of a residual signal in prediction in video codec, thereby minimizing a mode determination error and improving video quality. A method for determining a video prediction mode comprises the following steps of: performing inter prediction for input video data and generating an SAD(Sum of Absolute Difference) value, the absolute sum of residual data, obtained from a difference between an original block and a prediction block(310); performing intra prediction for the input video data by block and generating an SAD(Sum of Absolute Difference) value, the absolute sum of residual data, or motion vectors obtained from the difference between the original block and the prediction block(320); acquiring a cost value with the SAD of the residual data generated in the intra prediction and the characteristics of a residual signal(330); acquiring a cost value with the SAD of the residual data generated in the inter prediction, the characteristics of the residual signal, or the motion vectors(340); comparing the cost value of the intra prediction with the cost value of the inter prediction(350); and selecting the intra prediction or the inter prediction with the minimum cost value as an encoding mode for block encoding(360).

Description

Method and apparatus for determining image prediction mode {Method and apparatus for deciding video prediction mode}

1 is a block diagram of a codec device according to the present invention.

2 is a flowchart illustrating a method of determining an image prediction mode according to the present invention.

The present invention relates to a video codec (Encoder / Decoder) apparatus, and more particularly, to a method and apparatus for determining an image prediction mode for determining an intra or inter prediction mode based on a cost in consideration of characteristics of a residual signal.

In general, video codecs use various and complex technologies to dramatically improve compression and image quality. For this reason, video codecs have been in the spotlight as application technologies for digital multimedia broadcasting (DMB), DVD, and the like instead of the conventional video compression standard.

The video codec performs a prediction process on sample data in block units to obtain a prediction block composed of prediction samples, and obtains a residual block obtained from the difference between the original block and the prediction block. Compress video data by transform & quantize.

There are two types of prediction methods, intra prediction and inter prediction. The inter prediction method performs motion compensation / estimation with reference to a reference picture which has undergone encoding / decoding and deblocking filtering. In addition, the intra prediction scheme performs prediction using data of neighboring blocks already encoded in the current picture. Accordingly, the video data compressed through the prediction, transformation, and quantization processes is compressed once again through an entropy coding process to be a bitstream according to a compression standard.

In this case, a criterion for determining an intra or inter prediction mode in prediction in a video codec is using a cost corresponding to a minimum value of Sum of Absolute Difference (SAD) or Sum of Absolute Transformed Difference (SATD). SAD is obtained by the absolute sum of the residual data of the macro block. However, even though the properties of the residual blocks are different, an error may occur due to the properties of the residual blocks. For example, a problem arises where the residual cost in which only one pixel is present may be greater than the residual cost in which the residual signal is spread over all pixels.

An object of the present invention is to provide an image prediction mode determination method for determining a cost value in consideration of SAD and residual signal characteristics as a criterion for intra or inter prediction mode determination in prediction in a video codec.

An object of the present invention is to provide an image prediction mode determination apparatus for determining a cost value in consideration of SAD and residual signal characteristics as a criterion for intra or inter prediction mode determination.

In order to solve the above technical problem, the present invention provides a method of determining the image prediction mode of the encoder,

Performing inter prediction and intra prediction by inputting data in macroblock units;

In the inter prediction and intra prediction, a cost value considering residual signal characteristics together with SAD is obtained, and a prediction having a minimum cost value is selected as an encoding mode.

In order to solve the above other technical problem, the present invention provides an image prediction apparatus of an encoder,

An inter prediction unit for performing inter prediction on blocks of input data;

An intra prediction unit that performs intra prediction on a block-by-block basis for input data;

A cost value is calculated based on the absolute sum of the residual data and the characteristics of the residual signal generated during intra prediction of the intra predictor, and the cost reflects the absolute sum of the residual data, the characteristics of the residual signal and the MV during inter prediction of the inter predictor. And a mode determination unit for calculating the value, comparing the cost value of the intra prediction with the cost value of the inter prediction, and selecting a prediction having the minimum cost value as the encoding mode.

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

1 is a block diagram of a codec device according to the present invention.

The codec device of FIG. 1 generally includes an encoding device and a decoder unit 180.

The encoding apparatus includes a prediction unit 110, a DCT unit 102 and a quantization unit 104, and an entropy coding unit 130.

The prediction block 110 performs a prediction process on sample data in units of blocks to obtain a prediction block composed of prediction samples, and intra prediction and inter prediction. Do this. In addition, intra prediction and / or inter prediction are performed according to properties of pictures such as an I picture, a P picture, and a B picture.

When the prediction block 110 is described in detail, the intra prediction unit 116 may include pixels of blocks adjacent to the block to be predicted in the picture decoded through the inverse quantization unit 117 and the inverse DCT unit 118. An intra mode for performing block prediction using data is performed. In this case, the intra prediction unit 116 generates SAD corresponding to the absolute sum of the residual data using the residual block data obtained by the difference between the original block data and the predicted block data.

The inter prediction unit 112 performs block prediction of the current picture using a reference picture stored in the memory 113 through the deblocking filter 114. In other words, the inter prediction unit 12 performs prediction by using information between pictures. The inter prediction unit 112 includes a motion estimation block (hereinafter abbreviated as ME) 112b and a motion compensation block (hereinafter abbreviated as MC) 112a. The reference picture or the reconstructed picture is stored in the memory (DMA) 113. In this case, the inter prediction unit 112 generates a SAD corresponding to the absolute sum of the residual data by using the residual block data obtained by the difference between the original block data and the predicted block data.

The motion estimation unit 112b includes a separate internal memory 112c therein. While external memory is accessed via the bus, internal memory 112c does not need to be accessed via the bus, so there is no burden on the bus. Motion estimation (hereinafter abbreviated as ME) uses a predetermined measurement function to find a motion vector representing a difference in the moving positions of both macroblocks by searching for a macroblock most similar to a macroblock in a current frame in a previous frame. It is a process. As a representative example of the method of finding the most similar macroblock, the search range is determined and the macroblock is moved by one pixel within the range, and the similarity between the two macroblocks is calculated by a predetermined measurement method to find the most similar macroblock. There is this. An example of the predetermined measurement method is to take an absolute value of the difference between the corresponding pixel value of a macroblock in a current frame and a macroblock in a search area, and add the values to the macroblock when the value is the smallest. There is a way to decide. More specifically, the degree of similarity between the macro block of the past frame and the current frame is determined using a similarity value, that is, a matching reference value calculated using pixel values included in the macro blocks of the past and current frames. The matching reference value is calculated using a predetermined measurement function, and a measurement function may be a sum of absolute difference (SAD), a sum of absolute transformed difference (SATD), or a sum of squared difference (SSD).

The mode determiner 114 may include a cost value reflecting the characteristics of the SAD generated by the intra predictor 116 and the residual signal, and a cost reflecting the characteristics of the SAD and the residual signal generated by the inter predictor 112 and the MV. Compare the values to select the prediction mode with the lowest cost value. The residual signal may have characteristics such as non-zero residual signal number or residual signal variation or residual signal number above a certain threshold or residual signal directionality or patterns in the frequency domain of residual signals. .

1 again, the subtraction unit 101 generates a difference sample between the block data of the current picture and the block data output from the prediction unit 110.

The DCT unit 102 performs discrete cosine transform on the sample output from the subtraction unit 101.

The quantization unit 104 quantizes the discrete cosine transform coefficients output from the DCT unit 102.

The entropy coding block 130 encodes the image data quantized by the quantizer 104 according to the mode information and the motion vector generated by the mode determiner 114 and the motion compensator 112a. ) Is output to the decoder 180 together with the intra / inter mode information and the motion vector information.

Meanwhile, the decoder 180 receives a bit stream from an encoding device. The decoder 180 decodes the bitstream based on the intra / inter mode information and the motion vector information included in the received bitstream.

2 is a flowchart illustrating a method of determining an image prediction mode according to the present invention.

First, image data in macroblock units is input.

Subsequently, inter prediction is performed on a block-by-block basis for the input image data, and at this time, an SAD that is an absolute sum of residual data obtained from the difference between the original block and the prediction block is generated in the intra prediction process (step 310). In this case, the characteristic of the residual signal generated from the difference between the original block and the prediction block may include the non-zero residual signal number or the displacement value of the residual signal or the residual signal number above a predetermined threshold or the directionality of the residual signal or the frequency domain of the residual signals. Patterns on the top.

In addition, intra prediction is performed on a block-by-block basis for the input data, and in this process, SAD or motion vectors, which are absolute sums of residual data obtained from the difference between the original block and the prediction block, are generated in the inter prediction process (step 320). In this case, the characteristic of the residual signal generated from the difference between the original block and the prediction block may include the non-zero residual signal number or the displacement value of the residual signal or the residual signal number above a predetermined threshold or the directionality of the residual signal or the frequency domain of the residual signals. Patterns on the top.

Subsequently, a cost value is calculated based on the absolute sum (SAD) of residual data generated during intra prediction and characteristics of the residual signal (step 330). In other words, the cost value is obtained as in Equation 1.

The characteristics of the residual data block are then the non-zero residual signal number in the residual signals or the variation of the residual signal or the number of residual signals above a certain threshold or the directionality of the residual signals or patterns in the frequency domain of the residual signals.

In operation 340, the absolute value of the residual data generated during inter prediction and the characteristics or negative vectors of the residual signal are calculated.

Next, the cost value of the intra prediction and the cost value of the inter prediction are compared (step 350).

Subsequently, the intra prediction or the inter prediction having the minimum cost value is selected as the encoding mode for the block encoding by comparing the cost value of the intra prediction with the cost value of the inter prediction (step 360).

The present invention is not limited to the above-described embodiment, and of course, modifications may be made by those skilled in the art within the spirit of the present invention.

The present invention can also be embodied as computer readable codes on a computer readable recording medium. The computer-readable recording medium includes all kinds of recording devices in which data that can be read by a computer system is stored. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, hard disk, floppy disk, flash memory, optical data storage device, and also carrier waves (for example, transmission over the Internet). It also includes the implementation in the form of. The computer readable recording medium can also be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion.

As described above, according to the present invention, the video quality can be improved by minimizing the mode decision error by determining the cost value in consideration of the distribution or the property of the residual signal in the prediction in the video codec as a criterion for the prediction mode decision. Therefore, the present invention can expect high efficiency when determining the encoding mode for an image (scene change or fast moving scene, etc.).

Claims (5)

In the image prediction mode determination method of the encoder, Performing inter prediction and intra prediction by inputting data in macroblock units; And obtaining a cost value in consideration of residual signal characteristics together with SAD when performing the inter prediction and intra prediction, and selecting an inter prediction or intra prediction having a minimum cost value as an encoding mode. The method of claim 1, wherein the residual signal characteristic is A non-zero residual signal number or a displacement value of the residual signal or a residual signal number above a predetermined threshold or the direction of the residual signal or patterns in the frequency domain of the residual signals. The method of claim 1, wherein the encoding mode selection process is performed. Obtaining a cost value from the absolute sum of the residual data and the characteristics of the residual signal in the intra prediction; Obtaining a cost value based on the absolute sum of the residual data and the characteristics of the residual signal in the inter prediction; And comparing the cost value of the intra prediction and the cost value of the inter prediction to select a prediction having a minimum cost value as an encoding mode. The method of claim 3, wherein the cost value of the inter prediction is a process of adding motion vectors to the absolute sum of the residual data and the characteristics of the residual signal. In the image prediction device of the encoder, An inter prediction unit for performing inter prediction on blocks of input data; An intra prediction unit that performs intra prediction on a block-by-block basis for input data; A cost value is calculated based on the absolute sum of the residual data and the characteristics of the residual signal generated during intra prediction of the intra predictor, and the cost reflects the absolute sum of the residual data, the characteristic of the residual signal and the motion vector during inter prediction of the inter predictor. And a mode determination unit that calculates a value, compares the cost value of the intra prediction with the cost value of the inter prediction, and selects a prediction having a minimum cost value as an encoding mode.

Images