JP2012023611A - Dynamic image encoding device, dynamic image decoding device, dynamic image encoding method, and dynamic image decoding method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable high encoding efficiency to be acquired by adaptively switching a binarization method based on the fact that continuous occurrence of coefficients with small absolute values is highly provable.SOLUTION: A dynamic image encoding device and a dynamic image encoding method execute: an array conversion step of arranging a plurality of quantization coefficients arranged in one dimension by scanning a plurality of quantization coefficients arranged in two dimensions in a block to be encoded; a symbol encoding step of performing entropy encoding for a SIG bit, a LAST2 bit, and a LAST bit; and a quantization coefficient value encoding step of performing entropy encoding for non-zero quantization coefficient values in the block to be encoded.

Description

  The present invention relates to a moving image encoding apparatus and moving image encoding method for encoding a moving image with high efficiency, a moving image decoding apparatus and a moving image decoding method for decoding a moving image encoded with high efficiency, and It is about.

As a coding method of a moving image coding apparatus that performs high-efficiency coding of a moving image, that is, a variable-length coding method of transform coefficients in moving image coding having frequency conversion of different block sizes, for example, the following patents Document 1 discloses a context adaptive binary arithmetic coding method (hereinafter referred to as “CABAC method”).
Hereinafter, in the CABAC method, a method for encoding a bit indicating whether or not each coefficient in the block is a significant coefficient and a coefficient value of each significant coefficient will be described.

FIG. 19 is an explanatory diagram showing the coefficients in the block rearranged in a predetermined scanning order.
Here, it is assumed that a method of scanning in a zigzag order from a low frequency component to a high frequency component is used.
At this time, whether or not each coefficient is a significant coefficient is expressed by bits called SIG bits and LAST bits, and these bits are arithmetically encoded.
The SIG bit is “1” if it is a significant coefficient, and “0” if it is not a significant coefficient.
The LAST bit is assigned only to the significant coefficient, and is “1” when the significant coefficient is the last coefficient in the block, and is “0” when the significant coefficient is not the last coefficient.
The probability table when the SIG bit and the LAST bit are arithmetically encoded is determined according to the block size and type (for example, intra-screen prediction luminance signal block, inter-picture prediction luminance signal block, intra-screen prediction color difference signal block, inter-picture prediction color difference). Signal block), and for each block type, each coefficient position (position in scanning order).

According to Patent Document 2 below, the coefficient value of a significant coefficient is encoded as ABS, which is an absolute value, and SIGN bits, which are positive and negative signs.
As shown in FIG. 14, the ABS is indicated by a symbol of unary binarization (binary binarization) or a binary symbol having a prefix and a suffix, and this prefix depends on the value of the ABS. It consists of the number 1 and the suffix consists of a zero order exponent Golomb code.
The ABS binary symbol obtained in this way is subjected to arithmetic coding for each bit, and then the SIGN bit is arithmetically coded, whereby the frequency conversion block is encoded.

In general, it is known that the absolute value of a significant coefficient in a frequency conversion block tends to have a large low-frequency component and a small high-frequency component in moving image coding.
In particular, in a large frequency transform block, the number of significant coefficients increases, and therefore, there is a high probability that coefficients having a small absolute value will occur continuously in the high frequency component.

Japanese Patent No. 4230188 Japanese Patent No. 4313757

  Since the conventional moving image coding apparatus is configured as described above, in the CABAC system, there is a high probability that a coefficient having a small absolute value is continuously generated in the high frequency component in the frequency conversion block. The binarization method is fixed regardless of the frequency component. For this reason, there has been a problem that sufficient encoding efficiency may not be obtained.

The present invention has been made to solve the above-described problems, and adaptively switches the binarization method by utilizing the high probability that a coefficient having a small absolute value is continuously generated in a high-frequency component. It is an object of the present invention to obtain a moving picture coding apparatus and a moving picture coding method that can obtain high coding efficiency.
Another object of the present invention is to obtain a moving picture decoding apparatus and a moving picture decoding method that can be applied to the above moving picture encoding apparatus and moving picture encoding method.

  In the moving picture encoding apparatus according to the present invention, when the variable length encoding unit performs variable length encoding on the quantization coefficient output from the quantization unit, the variable length encoding unit is two-dimensionally arranged in the encoding target block. An array conversion step of scanning a plurality of quantization coefficients and arranging the plurality of quantization coefficients in a one-dimensional manner, and whether or not the quantization coefficients arranged in a one-dimensional manner in the array conversion step are non-zero When the first 1-bit symbol and the quantized coefficients arranged one-dimensionally in the array conversion step are non-zero, among the quantized coefficients arranged behind the quantized coefficients in the scanning order, Among the second 1-bit symbols indicating whether or not there is a quantized coefficient having an absolute value greater than a predetermined threshold, and the quantized coefficient arranged behind the quantized coefficient in the scanning order, the threshold Quantizer with larger absolute value If there is not, the entropy coding is performed on the third 1-bit symbol indicating whether or not there is a non-zero quantized coefficient among the quantized coefficients arranged behind the quantized coefficient in the scanning order. The symbol encoding step and the quantization coefficient value encoding step for entropy encoding the value of the non-zero quantization coefficient in the encoding target block are executed.

  According to the present invention, when the variable length coding means performs variable length coding on the quantization coefficient output from the quantization means, a plurality of quantization coefficients arranged two-dimensionally in the block to be coded. And a first 1 bit indicating whether or not the quantized coefficients arranged one-dimensionally in the array transformation step are non-zero. If the symbol and the quantized coefficient arranged one-dimensionally in the array conversion step are non-zero, the quantized coefficient arranged behind the quantized coefficient in the scanning order is larger than a predetermined threshold. Among the second 1-bit symbol indicating whether or not there is a quantized coefficient having an absolute value and the quantized coefficient arranged behind the quantized coefficient in the scanning order, an absolute value larger than the threshold is set. If there is no quantization coefficient to have, A symbol encoding step for entropy encoding a third 1-bit symbol indicating whether or not there is a non-zero quantized coefficient among the quantized coefficients arranged in order behind the quantized coefficient; Since the quantization coefficient value encoding step for entropy encoding the value of the non-zero quantization coefficient in the encoding target block is executed, the number of occurrences of the binary symbols in the high frequency component is suppressed. Thus, there is an effect that high encoding efficiency can be obtained.

It is a block diagram which shows the moving image encoder by Embodiment 1 of this invention. It is a flowchart which shows the processing content of the moving image encoder by Embodiment 1 of this invention. It is explanatory drawing which shows the example by which the largest encoding block is divided | segmented hierarchically into several encoding object block. (A) shows the distribution of partitions after division, and (b) is an explanatory diagram showing a situation where a coding mode m (B ⁿ ) is assigned by hierarchical division. It is explanatory drawing which shows the basic principle of the entropy encoding method of the quantization coefficient (transform coefficient block) which is the compression data in the variable length encoding part. It is explanatory drawing which shows an example of the number of SIG bits, a LAST2 bit, a LAST bit, and an ABS bit. It is explanatory drawing which shows an example of the number of SIG bits, a LAST2 bit, a LAST bit, and an ABS bit. It is explanatory drawing which shows an example of symbolization limited to "1" to "TH-1". It is explanatory drawing which shows an example of the binarization table by a Golomb code. It is explanatory drawing which shows an example of a block type, the value of a coefficient, and a category. It is a block diagram which shows the moving image decoding apparatus by Embodiment 1 of this invention. It is a flowchart which shows the processing content of the moving image decoding apparatus by Embodiment 1 of this invention. It is explanatory drawing which shows the basic principle of the entropy decoding method of the transform coefficient block in the variable length decoding part 31. FIG. It is explanatory drawing which shows the binarization symbol of ABS. It is explanatory drawing which shows the basic principle of the entropy encoding method of the transform coefficient block in the variable length encoding part 13 in Embodiment 2 of this invention. It is explanatory drawing which shows the basic principle of the entropy decoding method of the transform coefficient block in the variable length decoding part 31 of Embodiment 2 of this invention. It is explanatory drawing which shows the basic principle of the entropy encoding method of the transform coefficient block in the variable length encoding part 13 in Embodiment 2 of this invention. It is explanatory drawing which shows the basic principle of the entropy decoding method of the transform coefficient block in the variable length decoding part 31 of Embodiment 2 of this invention. It is explanatory drawing which shows what has rearranged the coefficient in a block in the predetermined scanning order.

Embodiment 1 FIG.
1 is a block diagram showing a moving picture coding apparatus according to Embodiment 1 of the present invention.
In FIG. 1, when a video signal indicating an input image is input, the block dividing unit 1 divides the input image into blocks having a coding block size determined by the coding control unit 2 (blocks of prediction processing units), A process of outputting an encoding target block that is a block of a prediction processing unit is performed. The block dividing unit 1 constitutes a block dividing unit.

The encoding control unit 2 determines the encoding block size and encodes the encoding target block output from the block dividing unit 1 from one or more selectable intra encoding modes and inter encoding modes. The process of determining the encoding mode with the highest is performed.
In addition, when the coding mode having the highest coding efficiency is the intra coding mode, the coding control unit 2 sets the intra prediction parameter used when performing the intra prediction process on the current block in the intra coding mode. When the coding mode having the highest coding efficiency is determined to be the inter coding mode, the inter prediction parameter used when performing the inter prediction process for the current block in the inter coding mode is executed. To do.
Further, the encoding control unit 2 performs a process of determining a prediction difference encoding parameter to be given to the transform / quantization unit 7 and the inverse quantization / inverse transform unit 8.

  If the coding mode determined by the coding control unit 2 is the intra coding mode, the changeover switch 3 outputs the block to be coded output from the block dividing unit 1 to the intra prediction unit 4, and the coding control unit If the coding mode determined by 2 is the inter coding mode, a process of outputting the coding target block output from the block division unit 1 to the motion compensation prediction unit 5 is performed.

The intra prediction unit 4 refers to the locally decoded image stored in the intra prediction memory 10 and uses the intra prediction parameter determined by the encoding control unit 2 to output the encoding target block output from the changeover switch 3. The intra prediction process is performed on the image to generate an intra predicted image (predicted image).
The motion compensated prediction unit 5 searches the motion vector by comparing the encoding target block output from the changeover switch 3 with the locally decoded image after filtering stored in the motion compensated prediction frame memory 12, and the motion vector Using the inter prediction parameters determined by the encoding control unit 2, an inter prediction process (motion compensation prediction process) is performed on the encoding target block to generate an inter predicted image (predicted image).
The changeover switch 3, the intra prediction unit 4, the motion compensation prediction unit 5, the intra prediction memory 10 and the motion compensation prediction frame memory 12 constitute a predicted image generation unit.

The subtraction unit 6 subtracts the intra prediction image generated by the intra prediction unit 4 or the inter prediction image generated by the motion compensation prediction unit 5 from the encoding target block output from the block division unit 1, A process of outputting a prediction difference signal (difference image) as a subtraction result to the transform / quantization unit 7 is performed.
The transform / quantization unit 7 refers to the prediction difference encoding parameter determined by the encoding control unit 2 and performs orthogonal transform processing (for example, DCT (discrete cosine transform)) on the prediction difference signal output from the subtraction unit 6. (Or orthogonal transform processing such as KL transform, in which a base design is made in advance for a specific learning sequence) is performed to calculate a transform coefficient, and the transform coefficient is quantized with reference to the prediction differential encoding parameter. And a process of outputting the compressed data (quantization coefficient of the difference image), which is the transformed transform coefficient, to the inverse quantization / inverse transform unit 8 and the variable length coding unit 13 is performed.
The subtracting unit 6 and the transform / quantization unit 7 constitute quantization means.

The inverse quantization / inverse transform unit 8 refers to the prediction difference encoding parameter determined by the encoding control unit 2 and inversely quantizes the compressed data output from the transform / quantization unit 7, and the prediction difference With reference to the encoding parameter, inverse orthogonal transform processing is performed on the transform coefficient that is compressed data after inverse quantization, and a local decoded prediction difference signal corresponding to the prediction difference signal output from the subtraction unit 6 is calculated. Implement the process.
The addition unit 9 includes the local decoded prediction difference signal calculated by the inverse quantization / inverse conversion unit 8, the intra prediction image generated by the intra prediction unit 4, or the inter prediction image generated by the motion compensation prediction unit 5. And a process of calculating a locally decoded image corresponding to the encoding target block output from the block dividing unit 1 is performed.

The intra prediction memory 10 is a recording medium that stores the locally decoded image calculated by the adding unit 9.
The loop filter unit 11 performs a predetermined filtering process on the local decoded image calculated by the adding unit 9 and performs a process of outputting the local decoded image after the filtering process.
The motion compensated prediction frame memory 12 is a recording medium that stores a locally decoded image after filtering processing.

The variable length coding unit 13 outputs the compressed data output from the transform / quantization unit 7, the output signal of the coding control unit 2 (coding mode, intra prediction parameter or inter prediction parameter, prediction difference coding parameter), The motion vector output from the motion compensation prediction unit 5 (when the encoding mode is the inter encoding mode) is subjected to variable length encoding to generate a bit stream.
However, when the variable length coding unit 13 performs variable length coding on the quantized coefficient that is the compressed data output from the transform / quantization unit 7, a plurality of two-dimensionally arranged multiple blocks are arranged in the coding target block. An SIG that scans a plurality of quantization coefficients and arranges the plurality of quantization coefficients in a one-dimensional manner and indicates whether or not the quantization coefficients that are arranged in the one-dimensional manner in the array transformation step are non-zero If the bit (first 1-bit symbol) and the quantization coefficient arranged one-dimensionally in the array conversion step are non-zero, the quantization coefficient arranged behind the quantization coefficient in the scanning order LAST 2 bits (second 1-bit symbol) indicating whether or not there is a quantized coefficient having an absolute value larger than a predetermined threshold, and the quantized coefficients arranged one-dimensionally in the array conversion step are not Zero and running If the absolute values of the quantized coefficients arranged in the order behind the quantized coefficients are all equal to or less than the predetermined threshold, the quantized coefficients arranged in the scan order after the quantized coefficients LAST bit (third 1-bit symbol) indicating whether or not there is a non-zero quantization coefficient, and a symbol encoding step for entropy encoding (SIG bit, LAST2 bit and LAST by symbol encoding step) (By encoding the bits entropy, the significance mapping indicating the position of the quantization coefficient having an absolute value larger than a predetermined threshold and the non-zero quantization coefficient in the current encoding block is encoded) And a quantization coefficient value encoding step for entropy encoding the value of the non-zero quantization coefficient in the encoding target block.
The variable length encoding unit 13 constitutes variable length encoding means.

In the example of FIG. 1, a block division unit 1, an encoding control unit 2, a changeover switch 3, an intra prediction unit 4, a motion compensation prediction unit 5, a subtraction unit 6, transform / quantization, which are components of the moving image encoding device. Unit 7, inverse quantization / inverse transform unit 8, addition unit 9, intra prediction memory 10, loop filter unit 11, motion compensated prediction frame memory 12, and variable length coding unit 13, each of which has dedicated hardware (for example, It is assumed that the CPU is configured by a semiconductor integrated circuit or a one-chip microcomputer). However, when the moving image encoding apparatus is configured by a computer, the block dividing unit 1, encoding control Unit 2, changeover switch 3, intra prediction unit 4, motion compensation prediction unit 5, subtraction unit 6, transform / quantization unit 7, inverse quantization / inverse transform unit 8, addition unit 9, loop filter unit 11, and variable length code Chemical unit 1 The processing contents stored programs describing the the memory of the computer, may execute a program that the CPU of the computer is stored in the memory.
FIG. 2 is a flowchart showing the processing contents of the moving picture coding apparatus according to Embodiment 1 of the present invention.

FIG. 11 is a block diagram showing a moving picture decoding apparatus according to Embodiment 1 of the present invention.
In FIG. 11, when a variable length decoding unit 31 receives a bit stream generated by a moving image encoding apparatus, compressed data, an encoding mode, and an intra prediction parameter (the encoding mode is an intra encoding mode) from the bit stream. Case), inter prediction parameters (when the coding mode is the inter coding mode), prediction differential coding parameters and motion vectors (when the coding mode is the inter coding mode) are subjected to variable length decoding. .
However, when the variable length decoding unit 31 performs variable length decoding of the quantized coefficients that are compressed data from the bitstream, the SIG bit (first) indicating whether or not the quantized coefficients arranged in one dimension are non-zero. LAST2 indicating whether there is a quantized coefficient having an absolute value larger than a predetermined threshold among the quantized coefficients arranged behind the non-zero quantized coefficient in the scanning order. LAST bit indicating whether or not there is a non-zero quantization coefficient among the bits (second 1-bit symbol) and quantization coefficients arranged behind the non-zero quantization coefficient in scanning order ( A symbol decoding step for entropy decoding the third 1-bit symbol), and the SIG bit, the LAST2 bit, and the LAST bit entropy decoded in the symbol decoding step A quantization coefficient value decoding step for specifying a position of a non-zero quantization coefficient in the decoding target block and entropy decoding a value of the non-zero quantization coefficient in the decoding target block; and in the decoding target block A plurality of quantization coefficients arranged in a one-dimensional manner are reversely scanned, and an array conversion step for arranging the plurality of quantization coefficients in a two-dimensional manner is executed.
The variable length decoding unit 31 constitutes a variable length decoding unit.

  The inverse quantization / inverse transform unit 32 refers to the prediction difference encoding parameter variable length decoded by the variable length decoding unit 31 and inversely quantizes the compressed data variable length decoded by the variable length decoding unit 31. A decoding prediction corresponding to the prediction difference signal output from the subtraction unit 6 of FIG. 1 is performed by referring to the prediction difference encoding parameter and performing an inverse orthogonal transform process on the transform coefficient that is the compressed data after the inverse quantization. A process for calculating a differential signal is performed. The inverse quantization / inverse transform unit 32 constitutes an inverse quantization means.

  The changeover switch 33 outputs the intra-prediction parameter variable-length decoded by the variable-length decoding unit 31 to the intra-prediction unit 34 if the coding mode variable-length decoded by the variable-length decoding unit 31 is the intra-coding mode. If the encoding mode variable-length decoded by the variable-length decoding unit 31 is an inter-coding mode, a process of outputting the inter prediction parameters and motion vectors variable-length decoded by the variable-length decoding unit 31 to the motion compensation unit 35 carry out.

The intra prediction unit 34 performs an intra prediction process on the decoding target block using the intra prediction parameter output from the changeover switch 33 while referring to the decoded image stored in the intra prediction memory 37, and performs the intra prediction process. A process of generating (predicted image) is performed.
The motion compensation unit 35 uses the motion vector and the inter prediction parameter output from the changeover switch 33 while referring to the decoded image after filtering stored in the motion compensated prediction frame memory 39, and performs inter prediction on the decoding target block. A process (motion compensation prediction process) is performed to generate an inter predicted image (predicted image).
The changeover switch 33, the intra prediction unit 34, the motion compensation unit 35, the intra prediction memory 37, and the motion compensated prediction frame memory 39 constitute a predicted image generation unit.

  The addition unit 36 adds the decoded prediction difference signal calculated by the inverse quantization / inverse conversion unit 32 and the intra prediction image generated by the intra prediction unit 34 or the inter prediction image generated by the motion compensation unit 35. Then, a process of calculating a decoded image corresponding to the encoding target block output from the block dividing unit 1 in FIG. 1 is performed.

The intra prediction memory 37 is a recording medium that stores the decoded image calculated by the addition unit 36.
The loop filter unit 38 performs a predetermined filtering process on the decoded image calculated by the adding unit 36 and performs a process of outputting the decoded image after the filtering process.
The motion compensated prediction frame memory 39 is a recording medium that stores the decoded image after the filtering process.

In the example of FIG. 11, the variable length decoding unit 31, the inverse quantization / inverse conversion unit 32, the changeover switch 33, the intra prediction unit 34, the motion compensation unit 35, the addition unit 36, and the intra prediction, which are components of the video decoding device. It is assumed that each of the memory 37, the loop filter unit 38, and the motion compensation prediction frame memory 39 is configured by dedicated hardware (for example, a semiconductor integrated circuit on which a CPU is mounted, a one-chip microcomputer, or the like). However, when the moving picture decoding apparatus is configured by a computer, the variable length decoding unit 31, the inverse quantization / inverse conversion unit 32, the changeover switch 33, the intra prediction unit 34, the motion compensation unit 35, the addition unit 36, and the loop A program describing the processing contents of the filter unit 38 is stored in the memory of a computer, and the CPU of the computer is stored in the memory. It is also possible to run the program.
FIG. 12 is a flowchart showing the processing contents of the moving picture decoding apparatus according to Embodiment 1 of the present invention.

Next, the operation will be described.
In the first embodiment, each frame image of a video is used as an input image, motion compensation prediction is performed between adjacent frames, and the obtained prediction difference signal is subjected to compression processing by orthogonal transformation / quantization, and then A moving picture coding apparatus that performs variable length coding to generate a bit stream and a moving picture decoding apparatus that decodes a bit stream output from the moving picture coding apparatus will be described.

The video encoding apparatus in FIG. 1 performs intra-frame and inter-frame adaptive encoding by dividing a video signal into regions of various sizes in response to local changes in the spatial and temporal directions of the video signal. It is characterized by that.
In general, a video signal has a characteristic that the complexity of the signal changes locally in space and time. When viewed spatially, on a certain video frame, for example, there are patterns with uniform signal characteristics in a relatively wide image area such as the sky and walls, etc., paintings including people and fine textures, etc. A pattern having a complicated texture pattern may be mixed in a small image area.
Even when viewed temporally, the change in the pattern of the sky and the wall locally in the time direction is small, but because the outline of the moving person or object moves rigidly or non-rigidly in time, the temporal change Is big.

In the encoding process, a prediction difference signal with small signal power and entropy is generated by temporal and spatial prediction to reduce the overall code amount. However, the parameters used for the prediction are set as large as possible in the image signal region. If it can be applied uniformly, the code amount of the parameter can be reduced.
On the other hand, if the same prediction parameter is applied to a large image region with respect to an image signal pattern having a large temporal and spatial change, a prediction error increases, so that the code amount of the prediction difference signal can be reduced. Can not.
Therefore, it is desirable to reduce the power and entropy of the prediction differential signal even in a region where the temporal and spatial changes are large even if the prediction target region is reduced and the amount of parameter data used for prediction is increased.

  In the first embodiment, in order to perform coding adapted to the general characteristics of such a video signal, first, prediction processing or the like is started from a predetermined maximum block size, and the video signal region is divided hierarchically. In addition, the prediction process and the encoding process of the prediction difference are adapted for each divided area.

The video signal format to be processed by the moving image encoding apparatus of FIG. 1 is a color video signal in an arbitrary color space such as a YUV signal composed of a luminance signal and two color difference signals, or an RGB signal output from a digital image sensor. In addition to the above, it is assumed that the video frame is an arbitrary video signal including a horizontal / vertical two-dimensional digital sample (pixel) sequence, such as a monochrome image signal or an infrared image signal.
However, the gradation of each pixel may be 8 bits, or a gradation such as 10 bits or 12 bits.

In the following description, for convenience, unless otherwise specified, it is assumed that the video signal of the input image is a YUV signal, and the two color difference components U and V are subsampled with respect to the luminance component Y 4: 2: 0. The case of handling format signals will be described.
A processing data unit corresponding to each frame of the video signal is referred to as a “picture”.
In the first embodiment, “picture” is described as a video frame signal that is sequentially scanned (progressive scan). However, when the video signal is an interlaced signal, “picture” is a unit constituting a video frame. It may be a field image signal.

First, the processing contents of the moving picture encoding apparatus in FIG. 1 will be described.
First, the encoding control unit 2 determines the size of the maximum encoding block used for encoding the picture to be encoded (current picture) and the upper limit of the number of hierarchies into which the maximum encoding block is divided (FIG. 2). Step ST1).
As a method of determining the size of the maximum coding block, for example, the same size may be determined for all the pictures according to the resolution of the video signal of the input image, or the local motion of the video signal of the input image The size difference may be quantified as a parameter, and a small size may be determined for a picture with high motion, while a large size may be determined for a picture with little motion.
As an example of how to determine the upper limit of the number of division layers, for example, when the motion of the video signal of the input image is intense, set the number of layers to be deeper so that a finer motion can be detected. There are methods such as setting to suppress the number of layers.

When the video signal of the input image is input, the block dividing unit 1 divides the picture into the picture of the input image with the maximum coding block size determined by the coding control unit 2, and outputs each divided picture.
In addition, the encoding control unit 2 divides each image area of the maximum encoding block size into encoding target blocks having the encoding block size hierarchically until reaching the upper limit of the number of division hierarchies previously determined. Then, the encoding mode for each encoding target block is determined (step ST2).

Here, FIG. 3 is an explanatory diagram showing an example in which the maximum coding block is hierarchically divided into a plurality of coding target blocks.
In FIG. 3, the maximum coding block is a coding target block whose luminance component indicated as “0th layer” has a size of (L ⁰ , M ⁰ ).
The encoding target block is obtained by hierarchically dividing the maximum encoding block as a starting point to a predetermined depth separately defined by a quadtree structure.
At the depth n, the encoding target block is an image area of size (L ⁿ , M ⁿ ).
However, L ⁿ and M ⁿ may be the same or different, but FIG. 3 shows the case of L ⁿ = M ⁿ .

Hereinafter, the encoding block size determined by the encoding control unit 2 is defined as the size (L ⁿ , M ⁿ ) in the luminance component of the encoding target block.
Since quadtree partitioning is performed, (L ^{n + 1} , M ^{n + 1} ) = (L ⁿ / 2, M ⁿ / 2) always holds.
Note that in a color video signal (4: 4: 4 format) in which all color components have the same number of samples, such as RGB signals, the size of all color components is (L ⁿ , M ⁿ ), but 4: 2. : When the 0 format is handled, the encoding block size of the corresponding color difference component is (L ⁿ / 2, M ⁿ / 2).

Later, represents the encoding target block of the n hierarchy B ^n, the encoding modes selectable by the encoding target block B ⁿ as represented by m (B ^n).
In the case of a color video signal composed of a plurality of color components, the encoding mode m (B ⁿ ) may be configured to use an individual mode for each color component, but hereinafter, unless otherwise specified, YUV The description will be made assuming that it indicates an encoding mode for a luminance component of an encoding block of a signal, 4: 2: 0 format.

The coding mode m (B ⁿ ) includes one or more intra coding modes (INTRA) and one or more inter coding modes (INTER). The coding control unit 2 The coding mode having the highest coding efficiency for the coding target block ^Bn is selected from all the coding modes available in the above or a subset thereof.

Furthermore, the encoding target block ^Bn is divided into one or a plurality of prediction processing units (partitions) by the block dividing unit 1 as shown in FIG.
Hereinafter, a partition belonging to the encoding target block B ⁿ is ^denoted as P _i ⁿ (i is a partition number in the nth layer).
How the partitioning of the encoding target block ^Bn is performed is included as information in the encoding mode m ( ^Bn ).
All partitions P _i ⁿ are subjected to prediction processing according to the coding mode m (B ⁿ ), but individual prediction parameters can be selected for each partition P _i ⁿ .

For example, the encoding control unit 2 generates a block division state as illustrated in FIG. 4 for the maximum encoding block, and identifies the encoding target block.
The shaded area in FIG. 4 (a) shows the distribution of the partitions after the division, and FIG. 4 (b) shows a situation in which the encoding mode m (B ⁿ ) is assigned by the hierarchical division in a quadtree graph. Yes.
Nodes surrounded by □ in FIG. 4B are nodes (encoding target blocks) to which the encoding mode m (B ⁿ ) is assigned.
For example, the encoding control unit 2 optimally selects D + λR (λ is a predetermined value) based on the code amount R when the block is divided and when the block is not divided and the image quality deterioration degree D. Appropriate hierarchy division can be performed by determining that it is the hierarchy division mode and repeating this for each hierarchy.

The changeover switch 3 is output from the block dividing unit 1 when the encoding mode m (B ⁿ ) determined by the encoding control unit 2 is an intra encoding mode (when m (B ⁿ ) ∈INTRA). The encoding target block ^Bn is output to the intra prediction unit 4.
On the other hand, when the encoding mode m (B ⁿ ) determined by the encoding control unit 2 is the inter encoding mode (when m (B ⁿ ) ∈INTER), the encoding target output from the block dividing unit 1 The block B ⁿ is output to the motion compensation prediction unit 5.

The intra prediction unit 4 has the coding mode m (B ⁿ ) determined by the coding control unit 2 in the intra coding mode (when m (B ⁿ ) ∈INTRA), and the block to be coded is switched from the changeover switch 3. When B ⁿ is received (step ST3), the encoding target block B is referred to by using the intra prediction parameter determined by the encoding control unit 2 while referring to the local decoded image stored in the intra prediction memory 10. and implementing intra prediction process for each partition _P ^{i n} in the ^n, it generates an intra prediction image _{P INTRAi} ⁿ (step ST4).
Since the image decoding apparatus needs to generate exactly the same intra prediction image and the intra prediction image P _INTRAi ^n, intra prediction parameters used for generating the intra prediction image P _INTRAi ⁿ is a variable from the encoding control unit 2 The data is output to the long encoding unit 13 and multiplexed into the bit stream.

The motion compensation prediction unit 5 is the encoding mode m (B ⁿ ) determined by the encoding control unit 2 is an inter coding mode (when m (B ⁿ ) ∈INTER), and the object to be encoded is selected from the changeover switch 3. When the block B ⁿ is received (step ST3), each partition P _i ⁿ in the encoding target block B ⁿ is compared with the locally decoded image after filtering stored in the motion compensated prediction frame memory 12, and the motion vector is compared. to explore its using the motion vector and coding inter-prediction parameters determined by the control unit 2, implemented inter prediction processing for each partition P _i ⁿ in the encoding target block B ^n, the inter prediction image generating a P _INTERi ⁿ (step ST5).
Since the image decoding apparatus must generate an identical inter prediction image and the inter-predicted image P _INTERi ^n, the inter prediction parameters used for generating the inter prediction image P _INTERi ^n, the variable from the encoding control unit 2 The data is output to the long encoding unit 13 and multiplexed into the bit stream.
In addition, the motion vector searched by the motion compensation prediction unit 5 is also output to the variable length encoding unit 13 and multiplexed into the bit stream.
For the prediction in the intra prediction unit 4 and the motion compensation prediction unit 5, for example, H.D. Any method such as H.264 intra prediction or motion compensation prediction is applicable to the present invention.

Subtraction unit 6, upon receiving the encoding target block ^{B n} from the block dividing unit 1 from its partition _P ^{i n} in the encoding target block ^{B n,} the intra prediction image _{P INTRAi} ⁿ generated by the intra prediction unit 4, or subtracts the inter prediction image _{P INTERi} ⁿ generated by the motion compensation prediction unit 5, and outputs the prediction difference signal _e ^{i n} a subtraction result to the transform and quantization unit 7 (step ST6).

When receiving the prediction difference signal e _i ⁿ from the subtraction unit 6, the transform / quantization unit 7 refers to the prediction difference encoding parameter determined by the encoding control unit 2 and is orthogonal to the prediction difference signal e _i ⁿ . Transformation processing (for example, DCT (discrete cosine transformation) or orthogonal transformation processing such as KL transformation in which a base design is made in advance for a specific learning sequence) is performed to calculate a transformation coefficient.
In addition, the transform / quantization unit 7 refers to the prediction difference encoding parameter, quantizes the transform coefficient, and performs the inverse quantization / inverse transform unit 8 and the variable length on the compressed data that is the transform coefficient after quantization. It outputs to the encoding part 13 (step ST7).

When receiving the compressed data from the transform / quantization unit 7, the inverse quantization / inverse transform unit 8 refers to the prediction difference encoding parameter determined by the encoding control unit 2 and dequantizes the compressed data. .
The inverse quantization / inverse transform unit 8 refers to the prediction differential encoding parameter and performs inverse orthogonal transform processing (for example, inverse DCT, inverse KL transform) on the transform coefficient that is the compressed data after inverse quantization. And a local decoded prediction difference signal corresponding to the prediction difference signal e _i ⁿ output from the subtraction unit 6 is calculated (step ST8).

Upon receiving the local decoded prediction difference signal from the inverse quantization / inverse transform unit 8, the adding unit 9 receives the local decoded prediction difference signal and the intra predicted image P _INTRAi ⁿ generated by the intra prediction unit 4 or motion compensation. by adding the inter prediction image P _INTERi ⁿ generated by the prediction unit 5, the local decoded partition image, or as a collection of the local decoded partition image, the encoding target block B ⁿ output from the block dividing unit 1 A corresponding local decoded image is calculated (step ST9).
The adding unit 9 outputs the locally decoded image to the loop filter unit 11 and stores the locally decoded image in the intra prediction memory 10.
This locally decoded image becomes an image signal for subsequent intra prediction.

When the loop filter unit 11 receives the local decoded image from the adding unit 9, the loop filter unit 11 performs a predetermined filtering process on the local decoded image and stores the local decoded image after the filtering process in the motion compensated prediction frame memory 12. (Step ST10).
The filtering process by the loop filter unit 11 may be performed in units of the maximum encoded block or individual encoded blocks of the input local decoded image, or a local decoded image corresponding to a macroblock for one screen is input. After being done, it may be performed for one screen at a time.

Variable length coding unit 13, the process of step ST3~ST9 of all the coding target block ^{B n} is completed (step ST11, ST12), and the compressed data output from the transform and quantization unit 7, the encoding control The encoding mode m (B ⁿ ) output from the unit 2 and the intra prediction parameter (when the encoding mode is the intra encoding mode) or the inter prediction parameter (the encoding mode is output from the encoding control unit 2). Bits indicating the encoding result obtained by variable-length encoding the inter-coding mode) and the motion vector output from the motion compensation prediction unit 5 (when the encoding mode is the inter-coding mode). A stream is generated (step ST13).

FIG. 5 is an explanatory diagram showing the basic principle of an entropy encoding method for quantized coefficients (transform coefficient blocks) that are compressed data in the variable length encoding unit 13.
First, a 1-bit symbol CBF (Coded_Block_Flag) is encoded for the transform coefficient block.
For example, the CBF bit can be configured to be “1” when a significant transformation coefficient exists in the block, and “0” when a significant transformation coefficient does not exist. Of course, this is not the case when the syntax of the upper layer indicates that the block has no significant conversion coefficient.
When the CBF bit is “0”, the encoding for the block is finished.

On the other hand, when the CBF bit indicates that a significant conversion coefficient exists, the following symbolization is performed.
First, a transform coefficient block in which significant transform coefficients are two-dimensionally arranged is converted into a one-dimensional array according to a predetermined scanning order (for example, zigzag scan).
Next, as shown in FIG. 6, information indicating the positions of significant conversion coefficients in the order of scanning is symbolized as SIG bits. For example, when there is a significant conversion coefficient, “1” is symbolized as the SIG bit, and when there is no significant conversion coefficient, “0” is symbolized as the SIG bit.

Only when “1” is converted into a symbol as the SIG bit, a value whose absolute value of the significant conversion coefficient is greater than or equal to a predetermined threshold value TH (for example, “2”) is set behind the position in the scanning order. A flag indicating whether or not it exists on the (high frequency component side) is symbolized as LAST2 bits.
That is, if there is no significant conversion coefficient having an absolute value equal to or greater than the threshold value TH after the position in the scanning order, “1” is symbolized, and if it exists, “0” is symbolized.
After the position where the LAST2 bit is “1”, a flag indicating whether or not a significant conversion coefficient is present behind the position (high frequency component side) in the scanning order is symbolized as a LAST bit.
That is, if there is no significant conversion coefficient after the position in the scanning order, “1” is symbolized, and if it exists, “0” is symbolized.
Note that once “1” is symbolized as the LAST2 bit, the symbolization of the LAST2 bit is completed, and only the LAST bit is symbolized.
In this way, a position where a significant conversion coefficient having an absolute value equal to or greater than the threshold TH exists in the scanning order and a position where a significant conversion coefficient having an absolute value whose absolute value is less than the threshold TH are specified.

Subsequently, the coefficient value of the significant conversion coefficient is symbolized. The coefficient value is composed of ABS, which is its absolute value, and SIGN bits, which are positive and negative signs. The SIGN bit is symbolized as “0” for a positive number and “1” for a negative number.
For the first significant coefficient in the scanning order, a value obtained by subtracting “1” from ABS is symbolized.

For the significant coefficient that is not the first in the scanning order, when the LAST2 bit is “0”, a value obtained by subtracting “1” from the ABS is symbolized, and when the LAST2 bit is “1”, the ABS is obtained. A value obtained by subtracting the threshold value TH from is symbolized.
When the LAST2 bit does not exist and only the LAST bit exists, a value obtained by subtracting “1” from the ABS is symbolized.
The ABS is configured to perform encoding in the reverse order of the scanning order of the SIG bit, the LAST2 bit, and the LAST bit, for example.

The symbolization of ABS is indicated by, for example, a symbol of unary binarization (unary binarization) or a binary symbol having a prefix and a suffix as shown in FIG. The suffix consists of a zero-order exponential Golomb code.
In the case where the LAST2 bit does not exist and only the LAST bit exists, since the maximum value of ABS is the threshold value TH-1 in the symbolization of ABS-1, as shown in FIG. Symbolization limited to TH-1 ″ is performed. As a special case, when TH = 2, it is self-evident that the ABS value is “1”, so there is no need to encode the ABS.
For example, when there is a sequence as shown in FIG. 6, when TH = 3, the present method can reduce the number of generated symbols by three compared to the conventional method. Further, when TH = 2, the number of symbols generated can be reduced by 7 in this method compared to the conventional method.

Entropy encoding is performed on each binarized symbol obtained in this manner, thereby encoding the frequency transform block.
As entropy coding here, arithmetic coding for binarized symbols may be performed, or variable length coding is performed by assigning a Huffman code to one or a plurality of binarized symbols. Also good.
When performing arithmetic coding or variable length coding on a binary symbol, high-efficiency coding can be performed by coding each symbol based on the occurrence probability.
By assigning a probability table to each binarized symbol and updating the probability table based on the actual occurrence history of symbols, the encoding efficiency can be further improved.
Furthermore, by switching the probability table according to the condition (context), it is possible to realize more efficient encoding. Hereinafter, an example of switching the probability table according to the context will be described.

In order to encode the SIG bit, the LAST2 bit, and the LAST bit, max_coeff-1 different probability tables are used for each block category (see FIG. 10).
The context number for specifying which probability table is used is always indicated by the corresponding coefficient scanning position of the coefficient.
The context number of the coefficient coeff [i] scanned as the i-th coefficient is
ctx_ID_SIG (coeff [i])
= Ctx_ID_LAST2 (coeff [i])
= Ctx_ID_LAST (coeff [i])
= I
It becomes.

In addition, two different models are used as context numbers in the probability table used for ABS coding. The first context number ctx_ID_ABS_1bin is applied to the first bit of the binarized symbol of ABS, and the other context number ctx_ID_ABS_rbins is applied to the remaining binarized symbols.
ctx_ID_ABS_1bin is assigned, for example, when a coefficient having ABS> 1 is encoded in a block, “4” is assigned as a context number, and when more than three coefficients of ABS = 1 are encoded in a block, context Assign 3 as the number.
In other cases, the number of ABS = 1 coefficients coded in the block is assigned as the context number.

Thus, by switching context according to the number of ABS = 1 coefficients and the rare number of ABS> 1, if the ABS> 1 coefficients are not encoded, the probability that ABS> 1 will occur is The value increases each time a significant coefficient is generated.
When the coefficient of ABS> 1 occurs once, the probability that the coefficient of ABS> 1 will be further increased.
For this reason, an appropriate probability table can be assigned according to conditions, and encoding can be performed with high encoding efficiency.

Also, ctx_ID_ABS_rbin is assigned, for example, as a context number “4” when more than four coefficients having ABS> 1 are encoded in the block, and in other cases, a coefficient having ABS> 1 is included in the block. Assign the number generated in step 1 as the context number.
In this way, by switching the context according to the number of ABS> 1 coefficients generated, the probability that the coefficient of ABS> 1 will be generated every time the coefficient of ABS> 1 is generated. A random probability table can be assigned, and encoding can be performed with high encoding efficiency.

Next, processing contents of the moving picture decoding apparatus in FIG. 11 will be specifically described.
When the variable length decoding unit 31 receives the bitstream generated by the moving picture encoding device in FIG. 1, the variable length decoding unit 31 performs variable length decoding processing on the bitstream (step ST21 in FIG. 12), and a picture of one frame or more. The frame size is decoded in sequence units composed of

That is, the variable length decoding unit 31 determines the maximum coding block size and the upper limit of the number of divided layers determined by the coding control unit 2 of the moving picture coding apparatus in FIG. 1 in the same procedure as the moving picture coding apparatus. (Step ST22).
For example, when the maximum encoding block size is determined in accordance with the resolution of the video signal, the maximum encoding block size is determined based on the decoded frame size in the same procedure as the moving image encoding apparatus.
When the maximum encoding block size and the upper limit of the number of division layers are multiplexed on the bit stream on the moving image encoding device side, values decoded from the bit stream are used.
As shown in FIG. 3, the moving image encoding apparatus is configured to encode an encoding mode and a conversion / conversion in units of encoding target blocks obtained by hierarchically dividing a maximum encoding block into a plurality of encoding target blocks starting from a starting point. The compressed data obtained by quantization is multiplexed into a bit stream.
The variable length decoding unit 31 that has received the bit stream decodes the division state of the maximum coding block included in the coding mode in the determined maximum coding block unit. Based on the decoded division state, a decoding target block (a block corresponding to the “encoding target block” of the moving image encoding apparatus in FIG. 1) is identified hierarchically (step ST23).

  Next, the variable length decoding unit 31 decodes the encoding mode assigned to the decoding target block. Based on the information included in the decoded encoding mode, the decoding target block is further divided into one or more prediction processing units, and the prediction parameters assigned to the prediction processing units are decoded.

  When the encoding mode assigned to the decoding target block (encoding target block) is the intra encoding mode, the variable length decoding unit 31 is one or more serving as a prediction processing unit included in the decoding target block. Intra prediction parameters are decoded for each partition.

The partition serving as the prediction processing unit is further divided into one or a plurality of partitions serving as the transform processing unit based on the transform block size information included in the prediction differential encoding parameter, and each partition serving as the transform processing unit. Then, the compressed data (transformed coefficient after quantization) is decoded (step ST24).
The variable length decoding unit 31 further decodes the residual signal component in each partition for each transform coefficient block.

FIG. 13 is an explanatory diagram showing the basic principle of the entropy decoding method for transform coefficient blocks in the variable length decoding unit 31.
First, the 1-bit symbol CBF is decoded for the transform coefficient block.
When the CBF bit is “1”, it indicates that a significant conversion coefficient exists in the block, and when the CBF bit is “0”, it indicates that no significant conversion coefficient exists in the block. Naturally, when the syntax of the upper layer already indicates that the block has no significant conversion coefficient, the decoding process below CBF is not performed.
When the CBF bit is “0”, the decoding for the block is finished, and the process proceeds assuming that all the transform coefficients in the transform block are “0”.

When the CBF bit is “1”, the following decoding process is performed.
First, information indicating the position of the significant conversion coefficient in the scanning order is decoded as SIG bits.
When the SIG bit is “1”, it indicates that there is a significant conversion coefficient, and when the SIG bit is “0”, it indicates that there is no significant conversion coefficient.
Only when “1” is decoded as the SIG bit, those whose absolute value of the significant conversion coefficient is greater than or equal to a predetermined threshold TH (for example, “2”) set in advance in the scanning order after the position ( The LAST2 bit, which is a flag indicating whether or not it exists on the high frequency component side), is decoded.
That is, when the LAST2 bit is “1”, it indicates that there is no significant conversion coefficient having an absolute value greater than or equal to the threshold value TH after the position in the scanning order, and when the LAST2 bit is “0”. Indicates that there is a significant conversion coefficient having an absolute value equal to or greater than the threshold value TH after the position in the scanning order.

After the position where the LAST2 bit is “1”, the LAST bit which is a flag indicating whether or not a significant conversion coefficient is present behind the position (high frequency component side) in the scanning order is decoded.
That is, when the LAST bit is “1”, it indicates that there is no significant conversion coefficient after the corresponding position in the scanning order, and when the LAST bit is “0”, the position is compared with the corresponding position in the scanning order. It shows that a significant conversion coefficient exists behind.
However, once “1” is decoded as the LAST2 bit, the decoding of the LAST2 bit is completed, and only the LAST bit is decoded.

In this manner, a position where a significant conversion coefficient having an absolute value equal to or greater than the threshold TH exists in the scanning order and a position where a significant conversion coefficient having an absolute value whose absolute value is less than the threshold TH are specified.
Subsequently, the coefficient value of the significant conversion coefficient is decoded.
The coefficient value is composed of ABS, which is its absolute value, and SIGN bits, which are positive and negative signs. The SIGN bit is decoded as “0” in the case of a positive number and “1” in the case of a negative number.

For the first significant coefficient in the scanning order, the value obtained by subtracting “1” from the ABS is decoded.
For the significant coefficient that is not the first in the scanning order, when the LAST2 bit is “0”, the value obtained by subtracting “1” from the ABS is decoded, and when the LAST2 bit is “1”, the threshold value TH is subtracted from the ABS. Values are decrypted.
When the LAST2 bit does not exist and only the LAST bit exists, a value obtained by subtracting “1” from the ABS is decoded.
The ABS is configured to perform decoding in the order opposite to the scanning order of the SIG bit, the LAST2 bit, and the LAST bit, for example.

For example, as shown in FIG. 14, the ABS is indicated by a symbol of unary binarization (binary binarization) or a binary symbol having a prefix and a suffix, and this prefix is added to the value of the ABS. It consists of a corresponding number of digits 1 and the suffix consists of a zero order exponent Golomb code.
In the case where the LAST2 bit does not exist and only the LAST bit exists, since the maximum value of ABS is the threshold value TH-1 in the decoding of ABS-1, as shown in FIG. Decoded as binary symbols limited to −1 ″. As a special case, when TH = 2, it is obvious that the ABS value is “1”, so there is no need to decode the ABS.

Each binarized symbol thus obtained is entropy-decoded, so that the frequency transform coefficients are decoded as a one-dimensional array.
Here, frequency conversion is performed by converting a one-dimensional array into a two-dimensional block by performing predetermined reverse scanning (for example, reverse zigzag scanning) which is reverse processing of scanning in the moving image encoding device of FIG. The block can be decoded.
However, the entropy decoding here needs to be a technique corresponding to the entropy coding in the moving picture coding apparatus of FIG.

In order to decode the SIG bit, the LAST2 bit, and the LAST bit, max_coeff−1 different probability tables are used for each block category (see FIG. 10).
The context number for specifying which probability table is used is always indicated by the corresponding coefficient scanning position of the coefficient.
The context number of the coefficient coeff [i] scanned as the i-th coefficient is
ctx_ID_SIG (coeff [i])
= Ctx_ID_LAST2 (coeff [i])
= Ctx_ID_LAST (coeff [i])
= I
It becomes.

Also, two different models are used as context numbers in the probability table used for ABS decoding. The first context number ctx_ID_ABS_1bin is applied to the first bit of the binarized symbol of ABS, and the other context number ctx_ID_ABS_rbins is applied to the remaining binarized symbols.
ctx_ID_ABS_1bin is assigned, for example, as a context number when a coefficient having ABS> 1 is decoded in a block, and as a context number when more than three coefficients of ABS = 1 are decoded in a block. 3 is assigned.
In other cases, the number of ABS = 1 coefficients decoded in the block is assigned as the context number.

In this way, by switching the context according to the generated ABS = 1 coefficient number and ABS> 1 rare number, the probability table selected at the time of encoding can be assigned according to the conditions, Can be decrypted.
Also, ctx_ID_ABS_rbin is assigned, for example, as a context number “4” when more than four coefficients having ABS> 1 are decoded in the block, and in other cases, a coefficient having ABS> 1 is set in the block. Assign the number of occurrences as the context number.
In this way, by switching the context according to the number of ABS> 1 coefficients generated, it is possible to assign the probability table selected at the time of encoding according to the conditions, and it is possible to suitably decode.

If the encoding mode m (B ⁿ ) variable-length decoded by the variable-length decoding unit 31 is an intra-encoding mode (when m (B ⁿ ) ∈INTRA), the changeover switch 33 is changed by the variable-length decoding unit 31. The variable length decoded intra prediction parameter is output to the intra prediction unit 34.
On the other hand, (the case of m ^{(B n)} ∈INTER) variable length decoded coding mode m ^{(B n)} is if the inter coding mode by the variable length decoding unit 31, variable length decoding by the variable length decoding unit 31 The inter prediction parameter and the motion vector thus output are output to the motion compensation unit 35.

The intra prediction unit 34 is an intra coding mode when the coding mode m (B ⁿ ) variable-length decoded by the variable length decoding unit 31 is m (B ⁿ ) ∈INTRA, and the intra prediction is performed from the changeover switch 33. When receiving the parameters (step ST25), the intra prediction parameters output from the changeover switch 33 while referring to the local decoded image stored in the intra prediction memory 37 in the same procedure as the intra prediction unit 4 in FIG. using, by implementing the intra prediction process for each partition _P ^{i n} the decoding target block ^{B n,} it generates an intra prediction image _{P INTRAi} ⁿ (step ST26).

The motion compensation unit 35 encodes from the changeover switch 33 when the encoding mode m (B ⁿ ) variable-length decoded by the variable-length decoding unit 31 is an inter encoding mode (when m (B ⁿ ) ∈INTER). When the target block ^Bn is received (step ST25), the motion vector output from the changeover switch 33 and the inter prediction parameter are used while referring to the decoded image after the filtering process stored in the motion compensated prediction frame memory 39. , by carrying out inter-prediction process for the decoding target block to generate an inter prediction image _{P INTERi} ⁿ (step ST27).

  When receiving the compressed data and the prediction difference encoding parameter from the variable length decoding unit 31, the inverse quantization / inverse conversion unit 32 performs the prediction difference encoding in the same procedure as the inverse quantization / inverse conversion unit 8 of FIG. With reference to the parameters, the compressed data is inversely quantized, and with reference to the prediction differential encoding parameter, the inverse orthogonal transform process is performed on the transform coefficient that is the compressed data after the inverse quantization. The decoded prediction difference signal corresponding to the prediction difference signal output from the subtraction unit 6 is calculated (step ST28).

Addition unit 36, decodes the prediction difference signal calculated by the inverse quantization and inverse transform unit 32, an intra prediction image P _{INTRAi n} generated by the intra prediction unit 34 ^or, inter prediction generated by the motion compensation unit 35 by adding the image P _INTERi ^n, as a collection of one or more of the decoded partition image included in the decoding target block, and outputs the decoded image to the loop filter unit 38, a memory 37 for intra prediction the decoded image (Step ST29).
This decoded image becomes an image signal for subsequent intra prediction.

The loop filter unit 11, the process of step ST23~ST29 for all current block ^{B n} is completed (step ST30), the output has been decoded image from the adder 36, and performs a predetermined filtering process, The decoded image after the filtering process is stored in the motion compensated prediction frame memory 39 (step ST31).
This decoded image becomes a reference image for motion compensation prediction and also becomes a reproduced image.

  As is apparent from the above, according to the first embodiment, when the variable length coding unit 13 of the moving picture coding apparatus performs variable length coding on the quantization coefficient output from the transform / quantization unit 7. , Scanning a plurality of quantized coefficients arranged two-dimensionally in the encoding target block, arranging the quantized coefficients one-dimensionally, and arranging the quantized coefficients one-dimensionally in the array-converting step SIG bit indicating whether or not the quantized coefficient is non-zero, and when the quantized coefficient arranged one-dimensionally in the array conversion step is non-zero, it is behind the quantized coefficient in the scanning order. LAST2 bits indicating whether or not there is a quantized coefficient having an absolute value larger than a predetermined threshold among the arranged quantized coefficients, and a quantized code arranged behind the quantized coefficient in scanning order Some of the coefficients are greater than the threshold If there is no quantized coefficient having an absolute value, an entropy of a LAST bit indicating whether or not there is a non-zero quantized coefficient among the quantized coefficients arranged behind the quantized coefficient in the scanning order is entropy. Since the configuration is such that the symbol encoding step for encoding and the quantization coefficient value encoding step for entropy encoding the value of the non-zero quantization coefficient in the encoding target block are performed, 2 in the high frequency component There is an effect that high coding efficiency can be obtained by suppressing the number of occurrences of the symbolized symbols.

Embodiment 2. FIG.
In the first embodiment, when the LAST2 bit or the LAST bit is encoded, the LAST2 bit or the LAST bit is encoded by 1 bit for the scanning position where the SIG bit is “1”.
However, especially when the frequency conversion block size is large, the number of consecutive “0” s as the LAST2 bit or the LAST bit increases.
Therefore, in the second embodiment, a contrivance is made to increase the coding efficiency when the number of consecutive “0” s as the LAST2 bit or the LAST bit increases.

FIG. 15 is an explanatory diagram showing the basic principle of the entropy coding method for transform coefficient blocks in the variable length coding unit 13 according to Embodiment 2 of the present invention.
Since other parts are the same as those in the moving picture coding apparatus according to the first embodiment, description thereof is omitted.
In the second embodiment, the variable length encoding unit 13 performs the encoding process of the frequency transform block, the number of “0” s in which LAST2 bits are continuous (NumCoeff2) and the number of “0” s in which LAST bits are continuous (NumCoeff). Is encoded to reduce the number of binarized symbols and obtain high encoding efficiency.

For example, by using the binarization table by Golomb code as shown in FIG. 9, binarization symbols are used, so that the conventional method of FIG. 6 has six symbols of LAST2 bits and LAST bits, TH = In the case of 3, 4 can be reduced, and in the case of TH = 2, 2 can be reduced.
Naturally, even if the binarization table is not a Golomb code, for example, it can be converted into a shorter symbol that is unary binarized, for example, a prefix that is unary binarized and a suffix that is Golomb code. The same effect can be obtained.

FIG. 16 is an explanatory diagram showing the basic principle of the entropy decoding method for transform coefficient blocks in the variable length decoding unit 31 according to the second embodiment of the present invention.
Similarly, in the variable length decoding unit 31 of the second embodiment, the LAST2 bit or the LAST bit is decoded as the number of consecutive “0”, for example, by decoding using the table shown in FIG. It is possible.
Here, the decoding is shown as the number of consecutive “0” s of LAST2 bits (NumCoeff2) and the number of consecutive “0s” of LAST bits (NumCoeff), but only one of them is consecutively “0”. The other one may be symbolized bit by bit with respect to the SIG bit as in the conventional case.

Also, as shown in FIGS. 17 and 18, encoding / decoding is performed with the number of consecutive LAST bits being “0” (NumCoeff) even in the conventional method using only the LAST bit without using the LAST2 bit. You may comprise so that it may perform.
In this case, the number of consecutive “0” s in the LAST bit is synonymous with encoding / decoding a value obtained by subtracting “1” from the number of significant coefficients in the block.
The assignment of “1” and “0” is only an example for the meanings of the CBF bit, SIG bit, LAST2 bit, LAST bit, and SIGN bit described in the first and second embodiments. The present invention is not limited to this, and the same effect can be obtained even if different assignments are made.
In addition, in the symbolization of ABS, a unary binarized symbol is used as a prefix and a zero-order exponent Golomb code is used as a suffix. However, the present invention is not limited to this. If the method assigns a long code, the effect of the present invention can be obtained.
The ABS encoding / decoding is configured to perform processing in the reverse order of the scanning order of the SIG bit, the LAST2 bit, and the LAST bit. Naturally, the processing is performed in the same order as the scanning order. May be.

  1 block division unit (block division unit), 2 encoding control unit, 3 changeover switch (prediction image generation unit), 4 intra prediction unit (prediction image generation unit), 5 motion compensation prediction unit (prediction image generation unit), 6 Subtraction unit (quantization unit), 7 transformation / quantization unit (quantization unit), 8 inverse quantization / inverse transformation unit, 9 addition unit, 10 intra prediction memory (prediction image generation unit), 11 loop filter unit, 12 motion-compensated prediction frame memory (predicted image generating means), 13 variable-length encoding section (variable-length encoding means), 31 variable-length decoding section (variable-length decoding means), 32 inverse quantization / inverse transform section (inverse quantum) 33) selector switch (predicted image generating means), 34 intra predictor (predicted image generating means), 35 motion compensator (predicted image generating means), 36 adder, 37 intra prediction memory (Predicted image generating means), 38 loop filter section, 39 motion compensated prediction frame memory (predicted image generating means).

Claims (6)

A block dividing unit that divides an input image into prediction processing unit blocks and outputs an encoding target block that is a prediction processing unit block; and a prediction process for the encoding target block that is output from the block dividing unit. A prediction image generating means for generating a prediction image, a prediction image generated by the prediction image generation means and a difference image between the encoding target blocks output from the block dividing means, and quantizing the difference image Quantizing means for outputting coefficients, and variable length coding means for generating a bitstream by variable length coding the quantized coefficients output from the quantization means,
The variable length coding means scans a plurality of quantization coefficients arranged two-dimensionally in the coding target block when variable length coding is performed on the quantization coefficient output from the quantization means. An array conversion step of arranging a plurality of quantization coefficients in a one-dimensional manner;
A first one-bit symbol indicating whether or not the quantized coefficients arranged one-dimensionally in the array conversion step are non-zero, and the quantized coefficients arranged one-dimensionally in the array conversion step are non-zero If there is, a second 1-bit symbol indicating whether or not there is a quantized coefficient having an absolute value greater than a predetermined threshold among the quantized coefficients arranged behind the quantized coefficient in the scanning order; If there is no quantization coefficient having an absolute value greater than the threshold value among the quantization coefficients arranged behind the quantization coefficient in the scanning order, the quantization coefficient is arranged behind the quantization coefficient in the scanning order. A symbol encoding step for entropy encoding a third 1-bit symbol indicating whether there are non-zero quantized coefficients among the quantized coefficients;
A moving picture coding apparatus that executes a quantization coefficient value coding step for entropy coding a value of a non-zero quantization coefficient in the coding target block.   The symbol encoding step entropy-encodes the first, second, and third 1-bit symbols using a probability table based on a context number that depends on a scanning position of each quantization coefficient. The moving picture coding apparatus according to 1. Performs variable-length decoding means for variable-length decoding quantized coefficients from a bitstream, inverse quantization means for inverse-quantizing quantized coefficients variable-length decoded by the variable-length decoding means, and prediction processing for a decoding target block Then, a predicted image generating unit that generates a predicted image, and an image addition that obtains a decoded image by adding the predicted image generated by the predicted image generating unit and the difference image indicated by the inverse quantization result of the inverse quantization unit Means and
The variable length decoding means includes a first 1-bit symbol indicating whether or not the one-dimensionally arranged quantized coefficients are non-zero when performing variable length decoding of the quantized coefficients from the bit stream, and scanning order. A second 1-bit symbol indicating whether or not there is a quantized coefficient having an absolute value greater than a predetermined threshold among the quantized coefficients arranged behind the non-zero quantized coefficient; A symbol decoding step for entropy decoding a third 1-bit symbol indicating whether or not there is a non-zero quantized coefficient among the quantized coefficients arranged behind the non-zero quantized coefficient in FIG.
Referring to the first, second, and third 1-bit symbols entropy-decoded in the symbol decoding step, the position of the non-zero quantized coefficient in the decoding target block is specified, and the non-decoding in the decoding target block A quantized coefficient value decoding step for entropy decoding the value of the quantized coefficient of zero;
A moving image characterized by performing reverse scanning of a plurality of quantized coefficients arranged one-dimensionally in the decoding target block and performing an array conversion step of arranging the plurality of quantized coefficients two-dimensionally Decoding device.   The symbol decoding step entropy-decodes the first, second, and third 1-bit symbols using a probability table based on a context number that depends on a scanning position of each quantized coefficient. Video decoding device. A block dividing unit divides the input image into blocks of prediction processing units, and outputs a block to be encoded that is a block of prediction processing units; and a prediction image generating unit includes the block dividing processing step A prediction image generation processing step for performing prediction processing on the output encoding target block and generating a prediction image; and a prediction unit generated by the quantization image in the prediction image generation processing step and the block division processing step A quantization processing step for quantizing the difference image of the block to be encoded output by step (a) and outputting a quantization coefficient for the difference image; and a quantization coefficient output by the variable length encoding means by the quantization processing step. Variable length coding processing step for generating a bit stream by variable length coding,
The variable length coding means scans a plurality of quantization coefficients arranged two-dimensionally in the coding target block when variable length coding is performed on the quantization coefficient output by the quantization processing step. An array conversion step of arranging a plurality of quantization coefficients in a one-dimensional manner;
A first one-bit symbol indicating whether or not the quantized coefficients arranged one-dimensionally in the array conversion step are non-zero, and the quantized coefficients arranged one-dimensionally in the array conversion step are non-zero If there is, a second 1-bit symbol indicating whether or not there is a quantized coefficient having an absolute value greater than a predetermined threshold among the quantized coefficients arranged behind the quantized coefficient in the scanning order; If there is no quantization coefficient having an absolute value greater than the threshold value among the quantization coefficients arranged behind the quantization coefficient in the scanning order, the quantization coefficient is arranged behind the quantization coefficient in the scanning order. A symbol encoding step for entropy encoding a third 1-bit symbol indicating whether there are non-zero quantized coefficients among the quantized coefficients;
A moving picture coding method comprising: performing a quantization coefficient value coding step for entropy coding a value of a non-zero quantization coefficient in the coding target block. The variable length decoding means performs variable length decoding on the quantization coefficient from the bit stream, and the inverse quantization means inversely quantizes the quantized coefficient variable length decoded in the variable length decoding processing step. The inverse quantization processing step, the prediction image generation means performs prediction processing on the decoding target block, and generates a prediction image, and the image addition means is generated in the prediction image generation processing step. An image addition processing step of adding a predicted image and a difference image indicated by the inverse quantization result in the inverse quantization processing step to obtain a decoded image,
The variable length decoding means includes a first 1-bit symbol indicating whether or not the one-dimensionally arranged quantized coefficients are non-zero when performing variable length decoding of the quantized coefficients from the bit stream, and scanning order A second 1-bit symbol indicating whether or not there is a quantized coefficient having an absolute value greater than a predetermined threshold among the quantized coefficients arranged behind the non-zero quantized coefficient; A symbol decoding step for entropy decoding a third 1-bit symbol indicating whether or not there is a non-zero quantized coefficient among the quantized coefficients arranged behind the non-zero quantized coefficient in FIG.
Referring to the first, second, and third 1-bit symbols entropy-decoded in the symbol decoding step, the position of the non-zero quantized coefficient in the decoding target block is specified, and the non-decoding in the decoding target block A quantized coefficient value decoding step for entropy decoding the value of the quantized coefficient of zero;
A moving image characterized by performing reverse scanning of a plurality of quantized coefficients arranged one-dimensionally in the decoding target block and performing an array conversion step of arranging the plurality of quantized coefficients two-dimensionally Decryption method.

Images