JP6704962B2 - Decoding device and decoding method - Google Patents

Description

The present invention relates to a process of decoding a block included in an image using motion compensation.

As a coding method for compressed recording of moving images, H.264 is used. H.264/MPEG-4 AVC (hereinafter abbreviated as H.264) is known (Non-Patent Document 1). H. In H.264, coding efficiency is improved by performing motion compensation by referring to other pictures as in the conventional coding method. The motion vector can be encoded in units of macroblocks (16 pixels×16 pixels) or blocks (8 pixels×8 pixels, etc.). In coding a motion vector, a predicted motion vector is calculated using the median value of motion vectors of surrounding blocks (left, upper, upper right), and the error between this and the motion vector to be coded is coded. In addition, H. In H.264, hierarchical coding can be performed. Non-Patent Document 1 is described in the chapter of Annex G Scalable video coding. In the case of spatial scalability, when the motion vector of the enhancement layer is encoded, the motion vector of the block of the base layer is expanded instead of the median value of the normal motion vector to predict the motion vector (hereinafter referred to as inter-layer motion vector predictor). Can be That is, if motion_prediction_flag is used and the value is true, the motion vector obtained by expansion can be used.

In addition, in recent years, H. As a successor to H.264, activities for international standardization of more efficient coding schemes have started. For this purpose, JCT-VC (Joint Collaborative Team on Video Coding) was established between ISO/IEC and ITU-T. Here, standardization is underway as a High Efficiency Video Coding coding system (hereinafter abbreviated as HEVC). In HEVC, Advanced Motion Vector Prediction (hereinafter, AMVP) has been proposed as a new motion vector encoding method (Non-Patent Document 2). In AMVP, not only the median value of the motion vectors of surrounding blocks is used as a reference motion vector, but the motion vector itself of surrounding blocks is also used as a reference motion vector. Further, not only the motion vector of the surrounding block but also the motion vector of the block at the same position of the previous picture in the coding order (hereinafter, the temporal motion vector predictor) is set as the motion vector predictor. The target motion vector is reduced by collecting motion vectors of the same component from these predicted motion vectors, and the closest motion vector is selected from them. A code for identifying the selected motion vector (hereinafter, prediction vector index code) and a prediction error as a result of prediction are encoded. This improves the coding efficiency.

ISO/IEC 14496-10:2010 Information technology---Coding of audio-visual objects--Part 10: Advanced Video Coding JCT-VC Contribution JCTVC-A124_r2. doc Internet <http://wftp3. itu. int/av-arch/jctvc-site/2010_04_A_Dresden/>

In HEVC, when an attempt is made to realize hierarchical coding, H.264/AVC is used. As in H.264, reference is made to the motion vector of the enhancement layer or the motion vector of the base layer. However, H. If the reference motion vector is selected like H.264, the above-mentioned method may be used, but the problem that the prediction vector index code by AMVP and the flag for selecting the prediction of the inter-layer motion vector are required and the coding efficiency is not improved. Is occurring.

Therefore, the present invention has been made to solve the above-mentioned problems, and an object of the present invention is to improve the encoding efficiency by performing the encoding of the motion vector in the enhancement layer of the hierarchical encoding with a simpler configuration. I am trying.

In order to solve the above problems, the decoding device of the present invention has the following configuration.

A decoding device for decoding at least one block included in an image from a bitstream hierarchically encoded using a plurality of layers,
A first vector that is included in an image that has been decoded before the image to which the decoding target block belongs in the first layer to which the decoding target image belongs and that is a motion vector of a block at a position corresponding to the decoding target block When,
A second vector which is included in an image in a second layer different from the first layer to which the image to be decoded belongs and which is a motion vector of a block at a position corresponding to the block to be decoded;
An acquisition unit that acquires a third vector that is included in the image in the first layer to which the image to be decoded belongs and that is a motion vector of a block around the block to be decoded,
From the first vector, the second vector, and the third vector, a candidate motion vector group to be acquired by the acquisition unit is determined, and among the candidate motion vector groups acquired based on the determination. Deciding means for deciding a motion vector corresponding to identification information indicating the order of the motion vector predictor decoded from the bitstream, as a motion vector predictor used for predicting the motion vector of the block to be decoded,
A quantized coefficient decoded from the bitstream, and a decoding unit that decodes the block to be decoded, based on the motion vector predictor determined by the determining unit,
The determining means performs a process of collecting motion vectors of the same component to determine the candidate vector group,
When the second vector exists in the candidate motion vector group, the determining means does not include the first vector as the candidate motion vector group, and the second vector and the third vector At least one of the above is determined as the candidate motion vector group ,
The decoding device, wherein the determining means arranges the second vectors in a predetermined order of the candidate motion vector group and determines the order of the other vectors based on the position of the block .

Furthermore, the decoding method of the present invention has the following configuration.

A decoding method for decoding a motion vector from a bitstream hierarchically encoded using a plurality of layers,
A first vector, which is a motion vector of a block at a position corresponding to the decoding target block in an image that has been decoded before the image to which the decoding target block belongs, in the first layer to which the decoding target image belongs;
A second vector which is a motion vector of a block at a position corresponding to the block to be decoded in the image in the second layer different from the first layer to which the image to be decoded belongs;
An acquisition step of acquiring a third vector, which is a motion vector of a block around the block to be decoded, of the image in the first layer to which the image to be decoded belongs,
Of the first vector, the second vector, and the third vector, a candidate vector group to be acquired by the acquisition step is determined, and a candidate vector group acquired based on the determination is determined, A determining step of determining a motion vector corresponding to identification information indicating the order of the motion vector predictor decoded from the bitstream, as a motion vector predictor used for predicting the motion vector of the block to be decoded;
A decoding step of decoding the block to be decoded, based on the quantized coefficient decoded from the bitstream and the motion vector predictor determined in the determining step,
In the determining step, a process of collecting motion vectors of the same component is performed to determine the vector group that is the candidate,
In the determining step, if the second vector exists in the candidate vector group, the first vector is not included as the candidate vector group, and the second vector and the third vector are not included. At least one of them is determined as the candidate vector group ,
In the determination step, decoding method the second vector are arranged in a predetermined order of the motion vector group to be the candidate, the order of the other vectors are characterized Rukoto is determined based on the position of the block.

According to the present invention, it is possible to provide a decoding device and method for processing data with improved coding efficiency.

FIG. 3 is a block diagram showing the configuration of the image encoding device according to the first embodiment. Detailed block diagram of the motion vector coding unit 102 in the image coding apparatus according to the first embodiment. Detailed block diagram of the motion vector encoding unit 105 in the image encoding device according to the first embodiment. Flowchart showing base layer motion vector coding processing in the image coding apparatus according to the first embodiment The flowchart which shows the enhancement layer motion vector coding process in the image coding apparatus which concerns on Embodiment 1. Diagram showing an example of blocks and motion vectors 3 is a block diagram showing the configuration of an image decoding device according to Embodiment 2. FIG. Detailed block diagram of the motion vector decoding unit 701 according to the second embodiment. Detailed block diagram of the motion vector decoding unit 704 according to the second embodiment. Flowchart showing base layer motion vector decoding processing in the image decoding apparatus according to the second embodiment Flowchart showing the enhancement layer motion vector decoding process in the image decoding device according to the second embodiment Detailed block diagram of the motion vector coding unit 105 in the image coding apparatus according to the third embodiment. The flowchart which shows the enhancement layer motion vector coding process in the image coding apparatus which concerns on Embodiment 3. Detailed block diagram of the motion vector decoding unit 704 according to the fourth embodiment. The flowchart which shows the enhancement layer motion vector coding process in the image coding apparatus which concerns on Embodiment 4. Diagram showing an example of blocks and motion vectors Block diagram showing a configuration example of hardware of a computer applicable to the image encoding device and the decoding device of the present invention

Hereinafter, the present invention will be described in detail based on its preferred embodiments with reference to the accompanying drawings. Note that the configurations shown in the following embodiments are merely examples, and the present invention is not limited to the illustrated configurations.

<Embodiment 1>
Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing an image coding apparatus according to this embodiment. In FIG. 1, 106 is a terminal for inputting image data. A frame memory 107 stores the input image data for each picture. A scaling unit 113 scales the input image data at a magnification of n/m (n and m are positive numbers). A frame memory 114 stores the scaled image data in units of pictures. Reference numerals 108 and 115 denote prediction units that cut out image data into a plurality of blocks, perform intra prediction that is intra-frame prediction, inter prediction by motion compensation, and the like to generate predicted image data. However, the prediction method is not limited to these. Further, a prediction error is calculated from the input image data and the predicted image data and output. When inter prediction is performed, another coded picture is referred to, a motion vector is calculated and output. Information necessary for prediction, for example, information such as intra prediction mode, is also output together with the prediction error. Reference numerals 109 and 116 denote transform/quantization units that orthogonally transform the prediction error in block units to obtain transform coefficients and further quantize them to obtain quantized coefficients. Reference numerals 112 and 119 are frame memories for storing reproduced image data. Reference numerals 111 and 118 denote image reproducing units. Quantization coefficients are input in block units, inverse quantization is performed to obtain transform coefficients, and further inverse orthogonal transformation is performed to reproduce prediction errors. The prediction units 108 and 115 generate predicted image data from the information about the motion vector and the prediction by appropriately referring to the frame memories 112 and 119, and generate and output reproduced image data from this and the reproduced prediction error. Reference numerals 110 and 117 are coefficient coding units that code the quantized coefficients output from the transform/quantization units 109 and 116 to generate quantized coefficient code data.

Reference numeral 100 is a motion vector encoding device according to the present invention. Reference numerals 101 and 104 denote motion vector holding units that hold motion vectors in block units. The motion vector holding unit 104 holds the motion vector generated by the prediction unit 108, and the motion vector holding unit 101 holds the motion vector generated by the prediction unit 115. Reference numerals 102 and 105 denote motion vector coding units that code motion vectors of the blocks generated by the prediction units 108 and 115 to generate motion vector code data. A motion vector scaling unit 103 scales a motion vector according to the scaling ratio of the scaling unit 113. A scaling unit 120 scales the image data in the frame memory 119 with a scaling ratio (m/n) opposite to that of the scaling unit 113.

121 and 122 are integrated encoding units. Code data of header information and information about prediction is generated, and the motion vector code data generated by the motion vector coding units 102 and 105 and the quantized coefficient code data generated by the coefficient coding units 110 and 117 are integrated. ..

Here, the frame memories 114 and 119, the prediction unit 115, the transform/quantization unit 116, the coefficient coding unit 117, the image reproduction unit 118, the motion vector holding unit 101, and the motion vector coding unit 102 convert the image data of the base layer. Encode. The frame memories 107 and 112, the prediction unit 108, the transform/quantization unit 109, the coefficient coding unit 110, the image reproduction unit 111, the motion vector holding unit 104, and the motion vector coding unit 105 code the image data of the enhancement layer. Turn into.

A multiplexing unit 123 multiplexes the coded data of the base layer and the coded data of the enhancement layer to form a bitstream. Reference numeral 124 denotes a terminal for outputting the bit stream generated by the multiplexing unit 123 to the outside.

The image encoding operation in the image encoding device will be described below. First, the coding operation of the base layer will be described.

The image data for one frame input from the terminal 106 is stored in the frame memory 107. The scaling unit 113 scales the image data stored in the frame memory 107 at a predetermined scaling factor n/m. In the case of spatial scalability, n/m has a value less than 1. The scaled image data is stored in the frame memory 114. The scaling factor is also input to the scaling unit 120 and the motion vector scaling unit 103.

The prediction unit 115 performs intra prediction or inter prediction by motion compensation. The scaling unit 120 scales the image data of the base layer of the frame memory 119 by m/n to generate reference image data of the base layer. In the case of inter prediction, the motion vector is calculated by referring to the coded picture stored in the frame memory 119 and the reference image data of the base layer. The calculated motion vector is input to the image reproducing unit 118 and the motion vector holding unit 101. The motion vector holding unit 101 holds a motion vector in block units. The prediction unit 115 also inputs the generated prediction error to the conversion/quantization unit 116. The quantized coefficient generated by the transform/quantization unit 116 is input to the coefficient coding unit 117 and entropy coded to generate quantized coefficient coded data. The entropy coding method is not particularly limited, but Golomb coding, arithmetic coding, Huffman coding, or the like can be used. The quantized coefficient, the motion vector, and the information about the prediction are input to the image reproduction unit 118, a reproduced image of the block to be encoded is generated, and stored in the frame memory 119.

The motion vector generated by the prediction unit 115 is coded by the motion vector coding unit 102 to generate motion vector coded data. FIG. 2 shows a detailed block diagram of the motion vector coding unit 102. In FIG. 2, 201 is a terminal to which a motion vector is input from the motion vector holding unit 101. Reference numeral 202 denotes a terminal, which inputs the motion vector of the coding target block from the prediction unit 115. A motion vector extraction unit 203 extracts a motion vector around a block to be coded or a motion vector of a previously coded picture through a terminal 201. A motion vector integration unit 204 collects motion vectors having the same component and generates a reference motion vector group. In the reference motion vector group, the grouped motion vectors are arranged in a predetermined order. The arrangement is not particularly limited, but it is arranged based on the order of component size, the order of the highest occurrence probability, the position of the extracted block, and the like. Any arrangement may be used as long as the arrangement is the same as that on the decoding side.

Reference numeral 205 denotes a motion vector predictor selection unit, which inputs the motion vector of the block to be encoded from the terminal 202 and the reference motion vector group from the motion vector integration unit 204. From the input reference motion vector group, the motion vector closest to the motion vector of the block to be encoded is selected as the motion vector predictor. An identification number for identifying the selected motion vector predictor with the reference motion vector group and the selected motion vector predictor are output. Reference numeral 206 denotes an index coding unit, which codes the output identification number to generate identification information code data. The coding method of the identification number is not particularly limited, but Golomb coding, arithmetic coding, Huffman coding, or the like can be used. A prediction unit 207 outputs a prediction error when the motion vector of the block to be encoded is predicted by the input motion vector predictor. A motion vector error encoding unit 208 encodes the motion vector prediction error to generate motion vector prediction error code data. The method for coding the motion vector prediction error is not particularly limited, but Golomb coding, arithmetic coding, Huffman coding, or the like can be used. A code forming unit 209 forms the identification vector code data and the motion vector prediction error code data to generate the motion vector code data. Reference numeral 210 denotes a terminal, which outputs the generated motion vector code data to the integrated coding unit 122 in FIG.

The coding operation of the motion vector of the base layer with the above configuration will be described below.

The motion vector extraction unit 203 inputs the motion vector of the block around the block of the base layer to be encoded and the temporal motion vector predictor via the terminal 201. These motion vectors are input to the motion vector integration unit 204, the motion vector components are compared, and the same components are collected to generate a reference motion vector group. FIG. 6 shows how motion vectors are extracted and integrated. Reference numeral 600 denotes a picture of the nth enhancement layer, and 601 denotes a picture of the n-1th enhancement layer. 603 is a block to be coded, and 604, 605, and 606 are blocks adjacent to the block to be coded. Reference numeral 607 denotes a picture of the (n-1)th enhancement layer, which represents a block at the same position as the block 603. Reference numeral 609 represents the motion vector of the encoding target block, and 610, 611, 612 and 613 represent the motion vector of each block. Here, when the motion vectors 611 and 612 have the same component, the reference motion vector group becomes the motion vectors 613, 612, and 610.

The generated reference motion vector group is input to the motion vector predictor selecting unit 205. The motion vector of the block to be encoded input from the terminal 202 is compared with the motion vector of the reference motion vector group in the motion vector predictor selecting unit 205. From the reference motion vector group, the motion vector of the component closest to the motion vector of the block to be encoded is selected as the predicted motion vector. An identification number indicating the order for identifying the selected motion vector with the reference motion vector group is generated and encoded by the index encoding unit 206 to generate identification information code data. The motion vector predictor is input to the prediction unit 207, the motion vector of the block to be coded is predicted, the prediction error is calculated, and the prediction error is output to the motion vector error coding unit 208. The motion vector error encoding unit 208 encodes the prediction error to generate motion vector prediction error code data. The identification information code data and the motion vector prediction error code data are shaped by the code shaping unit 209 and output from the terminal 210 as motion vector code data.

Returning to FIG. 1, the generated motion vector code data and the quantized coefficient code data generated by the coefficient coding unit 117 are input to the integrated coding unit 122 to generate code data in block units. The generated code data is input to the multiplexing unit 123.

Subsequently, the encoding operation of the image data of the enhancement layer will be described.

The prediction unit 108 divides the image data stored in the frame memory 107 into blocks, and performs intra prediction or inter prediction by motion compensation in block units. In the case of inter prediction, the motion vector is calculated by referring to the coded picture stored in the frame memory 112. The calculated motion vector is input to the image reproduction unit 111 and the motion vector holding unit 104. The motion vector holding unit 104 holds a motion vector in block units. The prediction unit 108 also inputs the generated prediction error to the conversion/quantization unit 109. The quantized coefficient generated by the transform/quantization unit 109 is input to the coefficient coding unit 110 and is entropy coded in the same manner as the coefficient coding unit 117 to generate quantized coefficient code data. Further, the quantization coefficient, the motion vector, and the information about the prediction are input to the image reproducing unit 111, the reproduced image of the block to be encoded is generated as in the image reproducing unit 118, and is stored in the frame memory 112.

The motion vector generated by the prediction unit 108 is coded by the motion vector coding unit 105 to generate motion vector coded data. Prior to the coding of the motion vector in the motion vector coding unit 105, the motion vector scaling unit 103 outputs the motion vector at the position of the base layer block corresponding to the coding target enhancement layer block from the motion vector holding unit 101. Extract. The extracted motion vector is multiplied by m/n according to the scaling factor (n/m) output from the scaling unit 113, and is output to the motion vector encoding unit 105 as an inter-layer predicted motion vector. FIG. 6 shows how motion vectors are extracted and integrated. Reference numeral 602 represents a picture of the corresponding base layer of the nth enhancement layer. Reference numeral 608 represents a block at a position corresponding to the block to be encoded in the enhancement layer. Reference numeral 614 represents a motion vector of the base layer block 608. The motion vector 614 is scaled to be an inter-layer predicted motion vector with the same precision as the resolution of the enhancement layer. Here, when the motion vectors 611 and 612 and the inter-layer predicted motion vector are the same component, the reference motion vector group is the inter-layer predicted motion vector and the motion vectors 613 and 610.

FIG. 3 shows a detailed block diagram of the motion vector coding unit 105. In FIG. 3, blocks that realize the same functions as the blocks in FIG. 2 are given the same numbers and their explanations are omitted. 301 is a terminal to which the inter-layer predicted motion vector is input from the motion vector scaling unit 103. A motion vector extraction unit 303 extracts a motion vector around a block to be coded in an enhancement layer or a motion vector of a previously coded picture through a terminal 201, and inputs an inter-layer predicted motion vector from a terminal 301. .. A motion vector integration unit 304 collects motion vectors of the same component and generates a reference motion vector group. In the reference motion vector group, the grouped motion vectors are arranged in a predetermined order. The arrangement method is not particularly limited, but the inter-layer predicted motion vector is arranged at the beginning. Other than that, they may be arranged based on the order of component size, the order of the highest occurrence probability, the position of the extracted block, and the like. Note that the terminal 201 is connected to the motion vector holding unit 104, the terminal 202 is connected to the prediction unit 108, and the terminal 210 is connected to the integrated coding unit 121.

The encoding operation of the motion vector of the enhancement layer with the above configuration will be described below.

The motion vector extraction unit 303 inputs a motion vector of a block around a block of a base layer to be encoded, a temporal direction motion vector predictor, and an inter-layer motion vector predictor. The motion vector integration unit 304 compares the motion vector components of these motion vectors and collects the same components to generate a reference motion vector group. In the reference motion vector group, the collected motion vectors are arranged in a predetermined order. The generated reference motion vector group is input to the motion vector predictor selecting unit 205. Hereinafter, similarly to the motion vector coding unit 102 of the base layer, the motion vector of the block to be coded is compared with the motion vector of the reference motion vector group, the predicted motion vector is selected, the identification number is generated, and the code is generated. To generate identification information code data. A motion vector prediction error of the block to be coded is calculated and coded to generate motion vector prediction error code data. The identification information code data and the motion vector prediction error code data are shaped by the code shaping unit 209 and output from the terminal 210 as motion vector code data.

Referring back to FIG. 1, the generated motion vector code data is input to the integrated coding unit 121, and the quantized coefficient code data generated by the coefficient coding unit 110 is input to the block unit code data. The generated code data is input to the multiplexing unit 123.

The multiplexing unit 123 multiplexes these code data according to a predetermined format, and outputs the coded data from the terminal 124 to the outside as a bit stream.

FIG. 4 is a flowchart showing the motion vector coding processing of the base layer in the image coding apparatus according to the first embodiment. First, in step S401, the motion vector of the block to be coded in the base layer is input and held for reference in the subsequent stage. In step S402, from among the motion vectors of the coded base layer in which the motion vectors of the blocks around the block to be coded of the base layer or the blocks at the same position in the coded picture of the base layer are held Extract. In step S403, the motion vectors of the same component are grouped by the motion vector extracted in step S402 and rearranged in a predetermined order to generate a reference motion vector group. In step S404, the reference motion vector group generated in step S403 is compared with the motion vector of the block to be coded in the base layer, the closest motion vector is selected as the motion vector predictor of the base layer, and identification information is generated. .. In step S405, the prediction error of the motion vector to be encoded is calculated from the motion vector predictor. In step S406, the identification information generated in step S404 is encoded. In step S407, the prediction error of the motion vector of the base layer is encoded.

Further, FIG. 5 is a flowchart showing the motion vector coding process of the enhancement layer in the image coding apparatus according to the first embodiment. 5, steps that realize the same functions as those of the blocks of FIG. 4 are assigned the same numbers, and explanations thereof are omitted. First, in step S501, a motion vector of a block to be coded in the enhancement layer is input and held for reference at a later stage. In step S502, the motion vector of the block around the block to be coded in the enhancement layer or the block at the same position in the picture of the coded enhancement layer is extracted from the held motion vectors of the enhancement layer. In step S503, the motion vector of the block of the base layer corresponding to the position of the block to be coded of the enhancement layer is extracted. In step S504, the motion vector of the extracted base layer is scaled to calculate an inter-layer predicted motion vector. In step S505, the motion vector extracted in step S502 and the motion vector of the same component in the inter-layer predicted motion vector are collected and rearranged in a predetermined order to generate a reference motion vector group. Hereinafter, similarly to the motion vector coding of the base layer in FIG. 4, in steps S504 to S506, a motion vector predictor is determined and identification information is generated. Further, a prediction error is calculated from the motion vector predictor. The prediction error of the generated identification information and motion vector is encoded.

With the above configuration and operation, in the motion vector coding of the enhancement layer, efficient coding can be performed using the motion vector of the base layer. When the scaling factor is 1, the scaling units 113 and 120 and the motion vector scaling unit 103 can be omitted.

Note that the motion vector integration unit 204 can always assign the inter-layer predicted motion vector to the head of the reference motion vector group. By this means, it is possible to easily assign a short code to an inter-layer motion vector predictor that is highly likely to be referenced, and thus it is possible to further improve coding efficiency.

Note that the motion vector extraction unit 203 inputs the motion vector of the block at the same position in the temporally-predicted motion vector encoded picture, but the present invention is not limited to this. For example, when the inter-layer motion vector predictor exists, the input of the temporal direction motion vector predictor may be omitted.

Although the motion vector integration unit 304 has described the case in which the inter-layer motion vector predictor is arranged at the head, the order following the temporal direction motion vector predictor may be used. Further, like the motion vector integration unit 204, it is not necessary to specifically determine the arrangement.

In addition, in order to improve the coding efficiency, the motion vectors having the same component are put together by the motion vector integration unit, but the present invention is not limited to this, and a fixed number of motion vectors may be selected. Furthermore, the motion vector integration unit may be omitted.

Further, the configuration may be such that the base layer coding apparatus and the enhancement layer coding apparatus are individually provided, and the scaling unit 113 and the scaling unit 120 are provided outside.

Although the block at the position shown in FIG. 6 is referred to as the motion vector of the reference motion vector group, the present invention is not limited to this. For example, a motion vector of a block adjacent to the right of the block 604 may be added, or a motion vector composed of the median of these motion vectors may be added. Further, the motion vector referenced from the base layer is not limited to the same position. As shown in FIG. 16, the motion vector 1604 of the block 1601 below the block 608, the motion vector 1605 of the lower right block 1602, and the motion vector 1606 of the right block 1603 may be referred to. Further, the motion vector of another block may be referred to.

<Embodiment 2>
FIG. 7 is a block diagram showing the configuration of the image decoding device according to the second embodiment of the present invention. In this embodiment, the decoding of the encoded data generated in the first embodiment will be described as an example.

Reference numeral 706 is a terminal for inputting an encoded bit stream. A separation unit 707 separates the bitstream into coded data of the base layer and coded data of the enhancement layer. The demultiplexing unit 707 performs the reverse operation of the multiplexing unit 123 in FIG. Integrated decoding units 708 and 715 decode the coded data of the header information and the prediction-related information for the base layer and the enhancement layer, separate the quantized coefficient coded data and the motion vector coded data, and output them to the subsequent stage. Reference numerals 709 and 716 denote coefficient decoding units, which perform operations reverse to those of the coefficient coding units 117 and 110 in FIG. 1 to reproduce quantized coefficients. Reference numerals 710 and 717 are inverse quantization/inverse transform units. With respect to the input quantized coefficient, the inverse operation of the transform/quantization units 116 and 109 of FIG. 1 is performed, the orthogonal transform coefficient is reproduced by the inverse quantization, and the prediction error is generated by the inverse orthogonal transform.

700 is a motion vector decoding device according to the present invention. Reference numerals 701 and 704 denote motion vector decoding units that decode the motion vector code data and reproduce the motion vector of the decoding target block. Reference numerals 702 and 705 denote motion vector holding units that hold motion vectors in block units. A motion vector scaling unit 706 scales the motion vector according to the same scaling ratio as the scaling unit 113 in FIG. The calculation of the scaling factor is not particularly limited. The ratio of the resolution of the base layer and the resolution of the enhancement layer may be calculated, or a predetermined value may be set.

Frame memories 712, 713, 719, and 720 store image data of reproduced pictures.

Reference numerals 711 and 718 denote image reproduction units, which input a motion vector, perform motion compensation with reference to image data of a decoded picture, and reproduce the image data using the decoded prediction error. A scaling unit 714 scales the image data of the base layer according to the same scaling ratio as the scaling unit 113 in FIG. The calculation of the scaling factor is not particularly limited. The ratio of the resolution of the base layer and the resolution of the enhancement layer may be calculated, or a predetermined value may be set. 721 and 722 are terminals, the terminal 721 outputs image data of the base layer, and the terminal 722 outputs image data of the enhancement layer.

Here, the integrated decoding unit 708, the coefficient decoding unit 709, the inverse quantization/inverse conversion unit 710, the motion vector decoding unit 701, the motion vector holding unit 702, the image reproduction unit 711, the frame memories 712 and 713 are the base layer code data. Is decoded and the image data of the base layer is reproduced. Further, the integrated decoding unit 715, the coefficient decoding unit 716, the inverse quantization/inverse conversion unit 717, the motion vector decoding unit 704, the motion vector holding unit 705, the image reproduction unit 718, the frame memories 719 and 720 store the encoded data of the enhancement layer. Decode and reproduce the image data of the enhancement layer.

The image decoding operation in the image decoding apparatus will be described below. In this embodiment, the bitstream generated in the first embodiment is decoded.

In FIG. 7, the bit stream input from the terminal 706 is input to the separation unit 707 and separated into the base layer code data and the enhancement layer code data, the former to the integrated decoding unit 708 and the latter to the integrated decoding unit 715. Is entered.

First, the decoding operation of base layer coded data will be described. The encoded data of the base layer is input to the integrated decoding unit 708. The integrated decoding unit 708 decodes the coded data of the header information and the information about the prediction, separates the quantized coefficient coded data and the motion vector coded data, inputs the motion vector coded data to the motion vector decoding unit 701, and quantizes the quantized coefficient. The coded data is input to the coefficient decoding unit 709.

The coefficient decoding unit 709 decodes the quantized coefficient code data and reproduces the quantized coefficient. The reproduced quantized coefficient is input to the inverse quantization/inverse transform unit 710, is inverse quantized to generate an orthogonal transform coefficient, and is further subjected to inverse orthogonal transform to reproduce the prediction error. The reproduced prediction error is input to the image reproduction unit 711.

The motion vector decoding unit 701 decodes the motion vector code data, and decodes the motion vector of the block to be decoded in the base layer. FIG. 8 shows a detailed block diagram of the motion vector decoding unit 701. In FIG. 8, reference numeral 801 denotes a terminal to which the motion vector code data of the decoding target block is input from the integrated decoding unit 708. A terminal 802 receives a motion vector from the motion vector holding unit 702. A motion vector extraction unit 803 extracts a motion vector around a block to be decoded in the base layer or a motion vector of a previously decoded picture through a terminal 802. Reference numeral 804 denotes a motion vector integration unit that performs the same operation as the motion vector integration unit 204 of FIG. 2 and collects motion vectors of the same component to generate a reference motion vector group. A code separation unit 805 separates the identification information code data and the motion vector prediction error code data from the motion vector code data. An index decoding unit 806 decodes the identification information code data and reproduces the identification information. The index decoding unit 806 performs the reverse operation of the index encoding unit 206 of FIG. 2 to perform decoding. A motion vector error decoding unit 807 decodes the motion vector prediction error code data and reproduces the motion vector prediction error of the block to be decoded. The motion vector error decoding unit 807 performs an operation reverse to that of the motion vector error encoding unit 208 in FIG. 2 to perform decoding. Reference numeral 808 is a motion vector predictor selection unit. A predicted motion vector is selected from the reference motion vector group using the reproduced identification information. A motion vector reproducing unit 809 reproduces the motion vector of the decoding target block of the base layer from the prediction error of the selected prediction motion vector and the reproduced motion vector. Reference numeral 810 denotes a terminal, which outputs the reproduced motion vector to the image reproducing unit 711 and the motion vector holding unit 702 in FIG. 7.

The decoding operation of the motion vector of the base layer with the above configuration will be described below. The motion vector code data of the base layer decoding target block input from the terminal 801 is input to the code separation unit 805 and is separated into identification information code data and motion vector prediction error code data. The former is input to the index decoding unit 806, and the latter is input to the motion vector error decoding unit 807. The index decoding unit 806 decodes the identification information code data and reproduces the identification information. Also, the motion vector error decoding unit 807 decodes the motion vector prediction error code data and reproduces the motion vector prediction error of the decoding target block of the base layer. Further, as shown in FIG. 6, the motion vector extraction unit 803 extracts the motion vector of the block around the block of the base layer to be decoded and the temporal prediction motion vector from the motion vector holding unit 702 via the terminal 802. The extracted motion vector is input to the motion vector integration unit 804, and the reference motion vector group is generated in the same manner as the motion vector integration unit 204 in FIG.

The motion vector predictor selection unit 808 inputs the identification information from the index decoding unit 806 and selects a motion vector predictor from the reference motion vector group. The selected motion vector predictor is input to the motion vector reproducing unit 809 together with the motion vector prediction error of the decoded target block of the reproduced base layer. The motion vector reproducing unit 809 reproduces the motion vector of the decoding target block from the reproduced prediction error and the predicted motion vector. The reproduced motion vector is output via the terminal 810.

Returning to FIG. 7, the reproduced motion vector is input to the image reproducing unit 711 and the motion vector holding unit 702. The motion vector holding unit 702 stores motion vectors in block units. The image reproduction unit 711 calculates the predicted image data of the decoding target block of the base layer from the frame memory 713 using the reproduced motion vector. Further, the prediction error reproduced by the inverse quantization/inverse transform unit 710 is input, and the image data of the base layer is reproduced using the predicted image data. The reproduced image data of the base layer is stored in the frame memory 712 and the frame memory 713 and used for reference. The reproduced image data of the base layer is output from the terminal 721.

Next, the decoding operation of the encoded data of the enhancement layer will be described. Prior to the decoding of the motion vector by the motion vector decoding unit 704, the motion vector scaling unit 703 extracts from the motion vector holding unit 702 the motion vector at the position of the base layer block corresponding to the decoding target enhancement layer block. The extracted motion vector is multiplied by m/n according to the scaling factor (n/m), and is output to the motion vector decoding unit 704 as an inter-layer predicted motion vector.

The encoded data of the enhancement layer is input to the integrated decoding unit 715. The integrated decoding unit 715 decodes the coded data of the header information and the information about the prediction, separates the quantized coefficient coded data and the motion vector coded data, inputs the motion vector coded data to the motion vector decoding unit 704, and quantizes the quantized coefficient. The coded data is input to the coefficient decoding unit 716.

The coefficient decoding unit 716 decodes the quantized coefficient code data and reproduces the quantized coefficient. The reproduced quantized coefficient is input to the inverse quantization/inverse transform unit 717, inversely quantized to generate an orthogonal transform coefficient, and further inverse orthogonal transform is performed to reproduce the prediction error. The reproduced prediction error is input to the image reproduction unit 718.

The motion vector decoding unit 704 decodes the motion vector code data, and decodes the motion vector of the block to be decoded in the enhancement layer. FIG. 9 shows a detailed block diagram of the motion vector decoding unit 704. In FIG. 9, blocks that realize the same functions as the blocks in FIG. 8 are given the same numbers, and explanations thereof are omitted. In FIG. 9, reference numeral 901 denotes a terminal, which inputs a motion vector from the motion vector scaling unit 703. Reference numeral 903 denotes a motion vector extraction unit, which inputs a motion vector around a block to be decoded in the enhancement layer or a motion vector of a previously coded picture as shown in FIG. 6 through a terminal 901. Also, the inter-layer motion vector predictor input from the terminal 901 is input. A motion vector integration unit 904 performs the same operation as that of the motion vector integration unit 904 of FIG. 8 and collects motion vectors of the same component to generate a reference motion vector group. In the reference motion vector group, the collected motion vectors are arranged in a predetermined order. The arrangement method is not particularly limited, but the inter-layer predicted motion vector is arranged at the beginning. Other than that, they may be arranged based on the order of component size, the order of the highest occurrence probability, the position of the extracted block, and the like. The terminal 801 is connected to the integrated decoding unit 715, the terminal 802 is connected to the motion vector holding unit 705, and the terminal 810 is connected to the motion vector holding unit 705 and the image reproducing unit 718.

The decoding operation of the motion vector of the enhancement layer with the above configuration will be described below. The motion vector extraction unit 903 inputs the motion vector of the block around the block of the enhancement layer to be decoded and the temporal prediction motion vector from the motion vector holding unit 705 via the terminal 802. Furthermore, the inter-layer motion vector predictor is input from the terminal 901. These motion vectors are input to the motion vector integration unit 904, and a reference motion vector group is generated in the same manner as the motion vector integration unit 804 in FIG. Hereinafter, similarly to the motion vector decoding unit 701 of the base layer, a motion vector predictor is selected from the identification information and the reference motion vector group. The motion vector of the decoding target block of the enhancement layer is reproduced according to the prediction error between the selected prediction motion vector and the motion vector of the reproduced decoding target block of the enhancement layer. The reproduced motion vector is output via the terminal 810.

Returning to FIG. 7, the reproduced motion vector is input to the image reproducing unit 718 and the motion vector holding unit 705. The motion vector holding unit 705 stores motion vectors in block units. The image reproducing unit 718 reads the corresponding image data from the frame memory 720 or the frame memory 713 using the reproduced motion vector, and inputs the image data scaled by the scaling unit 714. Predicted image data of the decoding target block of the enhancement layer is calculated from these image data. Further, the prediction error reproduced by the inverse quantization/inverse transform unit 717 is input, and the image data of the enhancement layer is reproduced using the predicted image data. The image data of the reproduced enhancement layer is stored in the frame memory 719 and the frame memory 720 and used for reference. The reproduced image data of the enhancement layer is output from the terminal 722.

FIG. 10 is a flowchart showing the motion vector decoding process of the base layer in the image decoding device according to the second embodiment. First, in step S1001, a motion vector of a block around a block to be decoded in the base layer and a temporal motion vector predictor are extracted from the held motion vector. In step S1002, the motion vectors of the same component are grouped by the motion vector extracted in step S1001 and rearranged in a predetermined order to generate a reference motion vector group. In step S1003, the identification information code data is decoded and the identification information is reproduced. In step S1004, a motion vector predictor is selected from the reference motion vector group generated in step S1002 based on the identification information reproduced in step S1003. In step S1005, the motion vector prediction error code data is decoded, and the motion vector prediction error of the block to be decoded in the base layer is reproduced. In step S1006, the motion vector of the decoding target block of the base layer is reproduced from the prediction error of the motion vector selected in step S1004 and the motion vector reproduced in step S1005. In step S1007, the reproduced motion vector is output and held for reference at a later stage.

Further, FIG. 11 is a flowchart showing motion vector decoding processing of the enhancement layer in the image encoding device according to the second embodiment. 11, steps that realize the same functions as those of the blocks of FIG. 10 are given the same numbers, and description thereof is omitted. First, in step S1101, a temporal motion vector predictor of a block or an enhancement layer around a block to be decoded of the enhancement layer is extracted from the retained decoded motion vectors of the enhancement layer. In step S1102, the motion vector of the block of the base layer corresponding to the position of the decoding target block of the enhancement layer is extracted. In step S1103, the motion vector of the extracted base layer is scaled to calculate an inter-layer predicted motion vector. In step S1104, the motion vector extracted in step S1102 and the motion vector of the same component in the inter-layer predicted motion vector are collected and rearranged in a predetermined order to generate a reference motion vector group. In step S1105, the identification information code data is decoded and the identification information is reproduced. In step S1106, the motion vector predictor is selected from the reference motion vector group generated in step S1104 based on the reproduced identification information. Hereinafter, similarly to the motion vector decoding of the base layer in FIG. 10, the motion vector prediction error code data is decoded and the motion vector prediction error is reproduced. The motion vector of the block to be decoded in the enhancement layer is reproduced from the selected motion vector predictor and the motion vector prediction error, and is output and held.

With the above configuration and operation, it is possible to decode a bitstream efficiently encoded using the motion vector of the base layer, which is generated in the first embodiment, and obtain a reproduced image. If the scaling factor is 1, the scaling unit 714 and the motion vector scaling unit 703 can be omitted.

Note that the motion vector integration unit 904 can always assign the inter-layer predicted motion vector to the head of the reference motion vector group. This makes it easier to assign a short code to the inter-layer motion vector predictor that is most likely to be referenced, and thus it is possible to further improve the coding efficiency.

Although the motion vector extraction unit 903 inputs the temporal motion vector predictor, the present invention is not limited to this. For example, when the inter-layer motion vector predictor exists, the input of the temporal direction motion vector predictor may be omitted.

Although the motion vector integration unit 904 has described the case where the inter-layer motion vector predictor is placed at the head, the motion vector integrating unit 904 may be arranged in the order following the temporal direction motion vector predictor. Further, like the motion vector integration unit 204, it is not necessary to specifically determine the arrangement.

Although the scaling factor is input from the motion vector scaling unit 703 in the present embodiment, the present invention is not limited to this, and the ratio of the resolution of the base layer and the resolution of the enhancement layer may be calculated individually, or a predetermined value may be calculated. May be set.

In addition, in order to improve the coding efficiency, the motion vectors having the same component are put together by the motion vector integration unit, but the present invention is not limited to this, and a fixed number of motion vectors may be selected. Furthermore, the motion vector integration unit may be omitted.

Alternatively, the decoding device for the base layer and the decoding device for the enhancement layer are individually provided, and the scaling unit 714 may be provided outside.

Although the block at the position shown in FIG. 6 is referred to as the motion vector of the reference motion vector group, the present invention is not limited to this. For example, a motion vector of a block adjacent to the right of the block 604 may be added, or a motion vector composed of the median of these motion vectors may be added. Further, the motion vector referenced from the base layer is not limited to the same position. As shown in FIG. 16, the motion vector 1604 of the block 1601 below the block 608, the motion vector 1605 of the lower right block 1602, and the motion vector 1606 of the right block 1603 may be referred to. Further, the motion vector of another block may be referred to.

<Embodiment 3>
In this embodiment, the image coding apparatus has the same configuration as that of the first embodiment shown in FIG. However, the configuration of the motion vector coding unit 105 of the enhancement layer is different. Therefore, the coding relating to the base layer and the coding of the quantized coefficient of the enhancement layer are the same as those in the first embodiment, and the description thereof will be omitted.

FIG. 12 is a block diagram showing a detailed configuration of the motion vector encoding unit 105 of this embodiment. In FIG. 12, parts having the same functions as those in FIG. 3 of the first embodiment are given the same numbers, and description thereof will be omitted. Reference numeral 1201 denotes a terminal, which inputs the inter-layer predicted motion vector from the motion vector scaling unit 103, similarly to the terminal 301 in FIG. A motion vector extraction unit 1203 extracts a motion vector around a block to be coded in the enhancement layer input from the terminal 201 or a motion vector of a previously coded picture. Reference numeral 1205 is a motion vector predictor selection unit. The motion vector of the block to be encoded in the enhancement layer is input from the terminal 202, the reference motion vector group is input from the motion vector integration unit 204, and the inter-layer predicted motion vector is input from the terminal 1201. From the input reference motion vector group, the motion vector closest to the motion vector of the block to be encoded is selected as the motion vector predictor.

The encoding operation of the motion vector of the enhancement layer with the above configuration will be described below.

The motion vector extraction unit 1203 extracts a motion vector around a block to be coded in the enhancement layer shown in FIG. 6 or a motion vector of a previously coded picture. The extracted motion vector is input to the motion vector integration unit 304. The motion vector integration unit 304 collects motion vectors having the same components in the input motion vectors as in the first embodiment, and generates a reference motion vector group. The motion vector predictor selecting unit 1205 selects a motion vector most similar to the motion vector of the current block to be coded as the motion vector predictor from the input reference motion vector group and inter-layer motion vector predictor. The identification information for identifying the selected motion vector predictor and the motion vector predictor are output.

Here, the difference from the first embodiment is that the inter-layer motion vector predictor is a candidate for selection independently of the reference motion vector group. Therefore, a motion vector having the same component may exist in the reference motion vector group. In this case, the motion vector predictor selecting unit 1205 selects the motion vector predictor between layers when the motion vector predictor between layers and the vector in the reference motion vector group can be selected as the motion vector predictor.

Hereinafter, the identification information is encoded by the index encoding unit 206 as in the first embodiment. Also, the prediction unit 207 calculates a prediction error from the motion vector of the current block and the motion vector predictor, and the motion vector error coding unit 208 codes the prediction error.

FIG. 13 is a flowchart showing motion vector coding processing of the enhancement layer in the image coding apparatus according to the third embodiment. In FIG. 13, steps that realize the same functions as those of the block in FIG. 4 are assigned the same numbers, and explanations thereof are omitted. First, steps S501 to S504 are the same as those in the first embodiment. That is, the motion vector around the block to be coded in the enhancement layer, or the motion vector of the previously coded picture and the inter-layer predicted motion vector are calculated. In step S1301, the motion vectors of the same components of the motion vector around the target block of the enhancement layer or the motion vector of the previously coded picture are collected and rearranged in a predetermined order to generate a reference motion vector group. To do. In step S1302, the inter-layer motion vector predictor is added to the head of the generated reference motion vector group. In step S404, a motion vector predictor is selected from the reference motion vector group as in the first embodiment. In step S1303, when the motion vector predictor is the same component as the inter-layer motion vector predictor, the process proceeds to step S1304. Otherwise, it proceeds to step S405. In step S1304, the inter-layer motion vector predictor is selected as the motion vector predictor. This prevents other motion vectors having the same component from being selected. Hereinafter, similarly to the motion vector coding process of the enhancement layer in FIG. 3 of the first embodiment, the calculation of the motion vector prediction error, the coding of the identification information, and the coding of the motion vector prediction error are performed in steps S405 to S407.

With the above configuration and operation, in the motion vector coding of the enhancement layer, the motion vector of the base layer can be used more preferentially than other motion vectors to perform efficient coding. This is because the inter-layer motion vector predictor has a very high correlation with the motion vector of the block of the enhancement layer.

Although the block at the position shown in FIG. 6 is referred to as the motion vector of the reference motion vector group, the present invention is not limited to this. For example, a motion vector of a block adjacent to the right of the block 604 may be added, or a motion vector composed of the median of these motion vectors may be added.

In addition, in order to improve the coding efficiency, the motion vectors having the same component are put together by the motion vector integration unit, but the present invention is not limited to this, and a fixed number of motion vectors may be selected. Furthermore, the motion vector integration unit may be omitted.

Although the block at the position shown in FIG. 6 is referred to as the motion vector of the reference motion vector group, the present invention is not limited to this.

<Embodiment 4>
In this embodiment, the image coding apparatus has the same configuration as that of FIG. 7 of the second embodiment. However, the configuration of the motion vector decoding unit 704 of the enhancement layer is different. Therefore, the decoding related to the base layer and the decoding of the quantized coefficient of the enhancement layer are the same as those in the second embodiment, and the description thereof will be omitted. In addition, in the present embodiment, the decoding of the encoded data generated in the third embodiment will be described as an example.

FIG. 14 is a block diagram showing a detailed configuration of the motion vector decoding unit 704 of this embodiment. In FIG. 14, parts that perform the same functions as those in FIG. 9 of the second embodiment are given the same numbers and description thereof is omitted. The terminal 1401 inputs the inter-layer predicted motion vector from the motion vector scaling unit 703, similarly to the terminal 901 in FIG. A motion vector extraction unit 1403 extracts the motion vector around the block to be coded in the enhancement layer shown in FIG. 6 or the motion vector of the previously coded picture, similarly to the motion vector extraction unit 1203 of the third embodiment. To do. A motion vector predictor selection unit 1405 selects a motion vector predictor from the motion vector predictor and the reference motion vector group according to the identification information.

The encoding operation of the motion vector of the enhancement layer with the above configuration will be described below.

The motion vector extraction unit 1403 extracts the motion vector around the decoding target block of the enhancement layer shown in FIG. 6 or the motion vector of the previously coded picture. The extracted motion vector is input to the motion vector integration unit 804. The motion vector integration unit 804 collects motion vectors having the same components in the input motion vectors as in the second embodiment, and generates a reference motion vector group. Further, as in the second embodiment, the index decoding unit 806 decodes the identification information coded data and reproduces the identification information. The motion vector predictor selecting unit 1405 selects a motion vector predictor from the inter-layer motion vector predictor or the reference motion vector group according to the identification information. Hereinafter, as in the second embodiment, the motion vector to be decoded in the enhancement layer is reproduced from the motion vector prediction error and the selected motion vector predictor.

FIG. 15 is a flowchart showing motion vector decoding processing of an enhancement layer in the image decoding device according to the fourth embodiment. 15, steps that realize the same functions as the blocks in FIG. 11 are assigned the same numbers, and explanations thereof are omitted. First, steps S1101 to S1103 are the same as those in the second embodiment. That is, the motion vector around the decoding target block of the enhancement layer, the motion vector of the previously decoded picture, and the inter-layer predicted motion vector are calculated. In step S1501, the motion vectors of the same component as the motion vector around the block to be decoded in the enhancement layer or the motion vector of the previously decoded picture are collected and rearranged in a predetermined order to generate a reference motion vector group. In step S1502, the inter-layer motion vector predictor is added to the head of the generated reference motion vector group. Hereinafter, similar to the motion vector decoding process of the enhancement layer of FIG. 11 of the second embodiment, the processes of steps S1105 to S1007 are performed. That is, the decoding of the identification information, the selection of the motion vector predictor, the reproduction of the motion vector prediction error, and the reproduction of the motion vector of the decoding target block of the enhancement layer are performed.

With the above configuration and operation, in the motion vector decoding of the enhancement layer, the motion vector of the base layer can be used more positively than other motion vectors to decode the coded data encoded with a smaller code amount. This is because the inter-layer motion vector predictor has a very high correlation with the motion vector of the block of the enhancement layer.

In addition, in order to improve the coding efficiency, the motion vectors having the same component are put together by the motion vector integration unit, but the present invention is not limited to this, and a fixed number of motion vectors may be selected. Furthermore, the motion vector integration unit may be omitted.

Although the block at the position shown in FIG. 6 is referred to as the motion vector of the reference motion vector group, the present invention is not limited to this.

<Embodiment 5>
The above-described embodiments have been described on the assumption that the processing units shown in FIGS. 1, 2, 3, 7, 8, 9, 12, and 14 are configured by hardware. However, the processing performed by each processing unit shown in these figures may be configured by a computer program.

FIG. 17 is a block diagram showing a configuration example of hardware of a computer applicable to the image display device according to each of the above embodiments.

The CPU 1701 controls the entire computer by using the computer programs and data stored in the RAM 1702 and the ROM 1703, and also executes the processes described above as those performed by the image processing apparatus according to each of the above embodiments. That is, the CPU 1701 functions as each processing unit shown in FIGS. 1, 2, 3, 7, 8, 9, 12, and 14.

The RAM 1702 has an area for temporarily storing a computer program or data loaded from the external storage device 1706, data acquired from the outside via an I/F (interface) 1709, and the like. Further, the RAM 1702 has a work area used when the CPU 1701 executes various processes. That is, the RAM 1702 can be allocated as a frame memory or can appropriately provide other various areas, for example.

The ROM 1703 stores the setting data of the computer, a boot program, and the like. The operation unit 1704 is composed of a keyboard, a mouse, and the like, and can be operated by the user of the computer to input various instructions to the CPU 1701. The display unit 1705 displays the processing result by the CPU 1701. The display unit 1705 is composed of, for example, a liquid crystal display.

The external storage device 1706 is a large capacity information storage device represented by a hard disk drive device. The external storage device 1706 is used to realize the OS (operating system) and the functions of the respective units shown in FIGS. 1, 2, 3, 7, 8, 9, 12, and 14 in the CPU 1701. Contains computer programs. Furthermore, each image data to be processed may be stored in the external storage device 1706.

The computer programs and data stored in the external storage device 1706 are appropriately loaded into the RAM 1702 under the control of the CPU 1701 and are processed by the CPU 1701. The I/F 1707 can be connected to a network such as a LAN and the Internet, and other devices such as a projection device and a display device. The computer acquires and sends various information via the I/F 1707. You can 1708 is a bus connecting the above-mentioned units.

The operation having the above-mentioned configuration is controlled by the CPU 1701 with the operation described in the above-mentioned flowchart being the center.

<Other embodiments>
The object of the present invention is also achieved by supplying a storage medium storing a code of a computer program for realizing the above-described functions to a system, and the system reading and executing the code of the computer program. In this case, the code itself of the computer program read from the storage medium realizes the functions of the above-described embodiments, and the storage medium storing the code of the computer program constitutes the present invention. Further, it includes a case where an operating system (OS) running on a computer performs a part or all of actual processing based on an instruction of the code of the program, and the processing realizes the above-described function. .

Further, it may be realized in the following forms. That is, the computer program code read from the storage medium is written in the memory provided in the function expansion card inserted in the computer or the function expansion unit connected to the computer. Then, based on the instructions of the code of the computer program, a case where the CPU included in the function expansion card or the function expansion unit performs a part or all of the actual processing to realize the above-described function is also included.

When the present invention is applied to the above storage medium, the storage medium stores the code of the computer program corresponding to the above-described flowchart.

Claims (2)

A decoding device for decoding at least one block included in an image from a bitstream hierarchically encoded using a plurality of layers,
A first vector that is included in an image that has been decoded before the image to which the decoding target block belongs in the first layer to which the decoding target image belongs and that is a motion vector of a block at a position corresponding to the decoding target block When,
A second vector which is included in an image in a second layer different from the first layer to which the image to be decoded belongs and which is a motion vector of a block at a position corresponding to the block to be decoded;
An acquisition unit that acquires a third vector that is included in the image in the first layer to which the image to be decoded belongs and that is a motion vector of a block around the block to be decoded,
From the first vector, the second vector, and the third vector, a candidate motion vector group to be acquired by the acquisition unit is determined, and among the candidate motion vector groups acquired based on the determination. Deciding means for deciding a motion vector corresponding to identification information indicating the order of the motion vector predictor decoded from the bitstream, as a motion vector predictor used for predicting the motion vector of the block to be decoded,
A quantized coefficient decoded from the bitstream, and a decoding unit that decodes the block to be decoded, based on the motion vector predictor determined by the determining unit,
The determining means performs a process of collecting motion vectors of the same component to determine the candidate vector group,
When the second vector exists in the candidate motion vector group, the determining means does not include the first vector as the candidate motion vector group, and the second vector and the third vector At least one of the above is determined as the candidate motion vector group ,
The decoding device, wherein the determining means arranges the second vectors in a predetermined order of the candidate motion vector group and determines the order of the other vectors based on the position of the block . A decoding method for decoding a motion vector from a bitstream hierarchically encoded using a plurality of layers,
A first vector, which is a motion vector of a block at a position corresponding to the decoding target block in an image that has been decoded before the image to which the decoding target block belongs, in the first layer to which the decoding target image belongs;
A second vector which is a motion vector of a block at a position corresponding to the block to be decoded in the image in the second layer different from the first layer to which the image to be decoded belongs
An acquisition step of acquiring a third vector, which is a motion vector of a block around the block to be decoded, of the image in the first layer to which the image to be decoded belongs,
Of the first vector, the second vector, and the third vector, a candidate vector group to be acquired by the acquisition step is determined, and a candidate vector group acquired based on the determination is determined, A determining step of determining a motion vector corresponding to identification information indicating the order of the motion vector predictor decoded from the bitstream, as a motion vector predictor used for predicting the motion vector of the block to be decoded;
A decoding step of decoding the block to be decoded, based on the quantized coefficient decoded from the bitstream and the motion vector predictor determined in the determining step,
In the determining step, a process of collecting motion vectors of the same component is performed to determine the vector group that is the candidate,
In the determining step, if the second vector exists in the candidate vector group, the first vector is not included as the candidate vector group, and the second vector and the third vector are not included. At least one of them is determined as the candidate vector group ,
The decoding method, wherein in the determining step, the second vectors are arranged in a predetermined order of the candidate motion vector group, and the order of the other vectors is determined based on the position of the block .

Images