JP2013207402A - Image encoding device and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce the amount of arithmetic operations while preventing a reduction in prediction accuracy when a peripheral candidate block is selected in motion vector prediction.SOLUTION: An image encoding device divides an image into encoding blocks and divides the encoding blocks into prediction blocks which are the objects of prediction in motion vector prediction as it encodes the image. The image encoding device comprises: a selection unit which selects an encoded block adjacent to a prediction block on the basis of the divided shape and position of prediction blocks in the encoding block; and an inter-prediction unit which applies motion compensation to the prediction block by using the motion vector of the encoded block selected by the selection unit.

Description

  The present invention relates to an image coding apparatus and a program for performing coding by dividing a video into block-like small areas.

  In recent years, with the advance of multimedia, there are many opportunities to handle moving images. Image data has a large amount of information, and a moving image, which is a set of still images, has an information amount exceeding 1 Gbit / sec in high-definition (1920 × 1080) broadcasting, which is the mainstream of television broadcasting.

  Therefore, the moving image data is H.264. H.264 / AVC (Advanced Video Coding) and standardization work HEVC (High Efficiency Video Coding) and the like are used for transmission and storage by compressing the amount of information.

  HEVC does not simply divide the screen into blocks that are encoding units from the upper left as in the conventional encoding technique, but newly adopts hierarchical division to enable block division of a plurality of hierarchies. A coding unit divided into a plurality of blocks is also called a CU (Coding Unit) or a coding block. For example, in HEVC, an image can be divided into 8 × 8, 16 × 16, 32 × 32, and the like and encoded.

  In HEVC, in order to efficiently express the prediction error signal, the CU can be divided into layers and divided into conversion units. The transform unit is also called a TU (Transform Unit) or transform block. For example, for a 32 × 32 CU, a TU with zero layer division is the same block as a 32 × 32 CU, and a TU with one layer division is a block divided into 16 × 16. In addition, a TU with two layer divisions is a block divided into 8 × 8, and a TU with three layer divisions is a block divided into 4 × 4.

  In HEVC, a CU is divided into prediction units that are a plurality of types of rectangular areas, and prediction processing is performed in each prediction unit. This prediction unit is also referred to as a PU (Prediction Unit) or a prediction block.

  FIG. 1 is a diagram illustrating an example of a prediction block used in inter-screen prediction or intra-screen prediction. As shown in FIG. 1, a CU can be divided into four types of PUs. For example, a 2N × 2N CU is divided into a 2N × 2N PU, an N × N PU, an N × 2N PU, and a 2N × N PU. There are various types of PUs depending on the size of the CU.

  Here, H. In a moving image compression coding system represented by H.264 and HEVC, compression by motion compensation using correlation between frames is performed. In order to deal with various motions in the screen, the entire image is hierarchically divided, and motion compensation is performed by block matching for each prediction block of the encoded block.

  In HEVC, in order to improve encoding performance, a motion vector prediction method for predicting a motion vector used for motion compensation from surrounding blocks is implemented in addition to a method of encoding a motion vector obtained by block matching as side information. A method with high performance is selected. In the motion vector prediction method, an adjacent encoded block group on the upper side or the left side of the encoding target block, or an encoded block existing at the same position in the time direction as the prediction block is used.

  An encoded block adjacent to the encoding target block is called an adjacent block, and a candidate used for motion vector prediction among the adjacent blocks is called a peripheral candidate block. In addition, a peripheral candidate block and a coded block existing at the same position as the coding target block in the time direction are referred to as a candidate block.

  Three blocks are selected from these candidate blocks, and the median value of the horizontal motion vectors is defined as the horizontal direction prediction value, and the median value of the vertical direction motion vectors is defined as the vertical direction prediction value.

  For example, in HEVC, a peripheral candidate block is selected from adjacent blocks. FIG. 2 is a diagram illustrating an example of a conventional peripheral candidate block. In the example illustrated in FIG. 2, the upper left adjacent block b111, the upper right adjacent blocks b112 and b113, and the lower left adjacent blocks b114 and b115 of the prediction block b101 are set as the peripheral candidate blocks. The adjacent blocks b121 to 124 are not defined as neighboring candidate blocks in HEVC. When the prediction block is located in the encoding target block other than the left end and the lower end, the lower left block of the prediction block can be used because it is an encoded block. When the prediction block is located at the upper end of the encoding target block, the block located at the upper right of the prediction block is an encoded block and can be used.

  In HEVC, there is a technique of selecting two candidate blocks from neighboring candidate blocks as shown in FIG. 2 and one candidate block at the same position in the time direction, and predicting the median value of these candidate blocks as a motion vector used for motion compensation. .

  In HEVC, in order to improve the encoding performance, the encoding performance is compared for all combinations with respect to the selection of these candidate blocks, and the combination of candidate blocks having the best performance is determined.

Hiroya Nakamura, `` JCTVC-F419: Unification of derivation process for merge mode and MVP, 14-22 July, 2011

  However, in the conventional technique, the adjacent block selected as the peripheral candidate block is at the same position with respect to the prediction block regardless of the shape or position of the prediction block. However, this excludes cases where there are adjacent blocks that cannot be referred to, such as the block to be encoded is a block at the end of the image. Therefore, the selected neighboring candidate block may include an adjacent block having a motion vector that is completely different from that of the prediction block.

  In this case, when motion vector prediction is performed using an adjacent block as a peripheral candidate block, the predicted motion vector is highly likely to be significantly different from the original motion vector of the prediction block. Hereinafter, a motion vector obtained by the motion vector prediction method is referred to as a predicted motion vector.

  When predictive motion vectors obtained by the median value for all combinations of peripheral candidate blocks are compared in terms of coding performance, there is a possibility that a peripheral candidate block having a motion different from the predicted block is selected as the optimal peripheral candidate block. Low. Therefore, there is a problem that the arithmetic processing increases without contributing to the encoding performance.

  Therefore, the present invention has been made in view of the above problems, and image coding that can reduce the amount of computation while preventing a decrease in prediction accuracy when selecting a peripheral candidate block in motion vector prediction. An object is to provide an apparatus and a program.

  An image encoding apparatus according to an aspect of the present invention divides an image into encoded blocks, divides the encoded block into prediction blocks to be predicted for motion vector prediction, and performs image encoding for encoding the image A selection unit that selects a coded block adjacent to the prediction block based on a division shape and a position of the prediction block in the coding block; and a coded block selected by the selection unit An inter prediction unit that performs motion compensation of the prediction block using the motion vector of the prediction block.

  In addition, when the encoding block is divided into the prediction blocks, the selection unit calculates an intersection between the division boundary line of the prediction block that is a current prediction target and the boundary line of the encoding block. A region dividing unit that extends the dividing boundary line from the intersection point toward the outer side of the coding block, and divides into two regions using the extended dividing boundary line, and among the two regions, And an exclusion unit that excludes the encoded block included in a region that does not include the current prediction target prediction block from selection candidates.

  The excluding unit may exclude an encoded block in which a part of the block or the center of gravity of the block is included in an area not including the current prediction target prediction block from the selection candidates.

  In addition, when the number of the selected encoded blocks is equal to or less than a predetermined value, the selection unit, among the encoded blocks included in an area that does not include the current prediction target prediction block, The number of encoded blocks may be selected in order from the closest prediction block to be predicted until a predetermined number is reached.

  Further, a program according to another aspect of the present invention divides an image into encoded blocks, divides the encoded block into prediction blocks to be predicted for motion vector prediction, and a computer that encodes the image, A selection step of selecting an encoded block adjacent to the prediction block based on a division shape and a position of the prediction block in the encoded block, and the prediction using a motion vector of the selected encoded block And an inter prediction step for performing block motion compensation.

  ADVANTAGE OF THE INVENTION According to this invention, when selecting a periphery candidate block by motion vector prediction, the amount of calculations can be reduced, preventing the fall of prediction accuracy.

The figure which shows an example of the prediction block used by inter prediction or intra prediction. The figure which shows an example of the conventional periphery candidate block. 1 is a block diagram illustrating an example of a schematic configuration of an image encoding device according to Embodiment 1. FIG. The block diagram which shows an example of a structure of a motion vector calculation part. The figure for demonstrating the 1st example of a periphery candidate block. The figure for demonstrating the 2nd example of a periphery candidate block. The figure for demonstrating the 3rd example of a periphery candidate block. The figure for demonstrating the estimated direction of a half line. The figure for demonstrating the addition selection of a periphery candidate block. The figure for demonstrating the parallel processing of a prediction block. The flowchart which shows an example of a motion vector prediction process. 5 is a flowchart illustrating an example of a process for selecting neighboring candidate blocks in the first embodiment. FIG. 6 is a block diagram illustrating an example of a configuration of an image processing apparatus according to a second embodiment.

  The present invention will be described below with reference to the accompanying drawings.

[Example 1]
<Configuration>
FIG. 3 is a block diagram illustrating an example of a schematic configuration of the image encoding device 10 according to the first embodiment. In the example illustrated in FIG. 3, the image encoding device 10 includes a preprocessing unit 100, a prediction error signal generation unit 101, an orthogonal transformation unit 102, a quantization unit 103, an entropy encoding unit 104, an inverse quantization unit 105, and an inverse orthogonal. Conversion unit 106, decoded image generation unit 107, deblocking filter unit 108, loop filter unit 109, decoded image storage unit 110, intra prediction unit 111, inter prediction unit 112, motion vector calculation unit 113, encoding control and header generation unit 114 and a predicted image selection unit 115. An outline of each part will be described below.

  The preprocessing unit 100 rearranges the pictures in accordance with the picture type, and sequentially outputs the picture type and the frame image for each frame. The preprocessing unit 100 also performs block division for encoding and the like.

  The prediction error signal generation unit 101 acquires block data obtained by dividing an encoding target image of input moving image data into blocks such as 32 × 32, 16 × 16, and 8 × 8 pixels.

  The prediction error signal generation unit 101 generates a prediction error signal based on the block data and the block data of the prediction image output from the prediction image selection unit 115. The prediction error signal generation unit 101 outputs the generated prediction error signal to the orthogonal transformation unit 102.

  The orthogonal transform unit 102 performs orthogonal transform processing on the input prediction error signal in transform units. The orthogonal transform unit 102 outputs the signal separated into horizontal and vertical frequency components by the orthogonal transform process to the quantization unit 103.

  The quantization unit 103 quantizes the output signal from the orthogonal transform unit 102. The quantization unit 103 reduces the code amount of the output signal by quantization, and outputs this output signal to the entropy encoding unit 104 and the inverse quantization unit 105.

  The entropy coding unit 104 entropy codes and outputs the output signal from the quantization unit 103, the motion vector information output from the motion vector calculation unit 113, the filter coefficient from the loop filter unit 109, and the like. Entropy coding is a method of assigning variable-length codes according to the appearance frequency of symbols.

  The inverse quantization unit 105 dequantizes the output signal from the quantization unit 103 and then outputs the output signal to the inverse orthogonal transform unit 106. The inverse orthogonal transform unit 106 performs an inverse orthogonal transform process on the output signal from the inverse quantization unit 105 and then outputs the output signal to the decoded image generation unit 107. By performing decoding processing by the inverse quantization unit 105 and the inverse orthogonal transform unit 106, a signal having the same level as the prediction error signal before encoding is obtained.

  The decoded image generation unit 107 adds the block data of the image subjected to motion compensation by the inter prediction unit 112 and the prediction error signal decoded by the inverse quantization unit 105 and the inverse orthogonal transform unit 106. The decoded image generation unit 107 outputs the decoded image block data generated by the addition to the deblocking filter unit 108.

  The deblocking filter unit 108 applies a filter for reducing block distortion to the decoded image output from the decoded image generation unit 107 and outputs the filtered image to the loop filter unit 109.

  The loop filter unit 109 is, for example, an ALF (Adaptive Loop Filter), and performs a filtering process using a decoded image subjected to the deblocking filtering process and an input image. For example, a Wiener filter is used for the filter processing, but other filters may be used.

  In addition, the loop filter unit 109 divides the input image into groups for each predetermined size, and generates an appropriate filter coefficient for each group. The loop filter unit 109 divides the decoded image subjected to the deblocking filter processing into groups for each predetermined size, and performs filter processing for each group using the generated filter coefficients. The loop filter unit 109 outputs the filter processing result to the decoded image storage unit 110 and accumulates it as a reference image. The predetermined size is, for example, an orthogonal transformation size.

  The decoded image storage unit 110 stores the input block data of the decoded image as new reference image data, and outputs the data to the intra prediction unit 111, the inter prediction unit 112, and the motion vector calculation unit 113.

  The intra prediction unit 111 generates a predicted image from the already-encoded reference pixels for the processing target block of the encoding target image. The intra prediction unit 111 performs prediction using a plurality of prediction directions and determines an optimal prediction direction.

  The inter prediction unit 112 performs motion compensation on the reference image data acquired from the decoded image storage unit 110 with the motion vector provided from the motion vector calculation unit 113. Thereby, block data as a motion-compensated reference image is generated.

  In addition, the inter prediction unit 112 may perform motion compensation using a prediction motion vector generated based on a motion vector included in a candidate block surrounding the prediction block in order to reduce the amount of calculation.

  The motion vector calculation unit 113 obtains a motion vector using the block data in the encoding target image and the reference image acquired from the decoded image storage unit 110. The motion vector is a value indicating a spatial deviation in units of blocks obtained using a block matching technique for searching for a position most similar to the processing target block from the reference image in units of blocks.

  The motion vector calculation unit 113 outputs the obtained motion vector to the inter prediction unit 112, and outputs motion vector information including information indicating the motion vector and the reference image to the entropy coding unit 104. In addition, the motion vector calculation unit 113 has a function of selecting neighboring candidate blocks from adjacent blocks of the prediction block. This function will be described later with reference to FIG.

  The block data output from the intra prediction unit 111 and the inter prediction unit 112 are input to the predicted image selection unit 115.

  The predicted image selection unit 115 selects one of the block data acquired from the intra prediction unit 111 and the inter prediction unit 112 as a predicted image. The selected prediction image is output to the prediction error signal generation unit 101.

  The encoding control and header generation unit 114 performs overall encoding control and header generation. The encoding control and header generation unit 114 notifies the intra prediction unit 111 of the presence / absence of slice division, the deblocking filter unit 108 notifies the deblocking filter presence / absence, and the motion vector calculation unit 113 provides a reference image. Notification of restrictions, etc.

  The encoding control and header generation unit 114 generates header information using the control result. The generated header information is output to the entropy encoding unit 104, and is output as a stream together with image data and motion vector data.

  The above-described configuration of the image encoding device 10 is merely an example, and the configuration may be changed depending on the manner of mounting. For example, the inter prediction unit 112 and the motion vector calculation unit 113 may have one configuration.

<Configuration for motion vector prediction>
Next, the configuration of the motion vector calculation unit 113 will be described. FIG. 4 is a block diagram illustrating an example of the configuration of the motion vector calculation unit 113. The motion vector calculation unit 113 illustrated in FIG. 4 includes a selection unit 201 for selecting neighboring candidate blocks.

  The motion vector calculation unit 113 has a configuration for calculating a general motion vector in addition to the configuration shown in FIG. Note that the inter prediction unit 112 may include the selection unit 201 that selects the peripheral candidate blocks used for determining the motion vector predictor.

  The selection unit 201 has a function of selecting neighboring candidate blocks from adjacent blocks of the prediction block. For example, the selection unit 201 selects a peripheral candidate block from neighboring blocks adjacent to the prediction block based on the division shape and position of the prediction block.

  In order to select a peripheral candidate block, the selection unit 201 includes, for example, an intersection calculation unit 211, a region division unit 212, an exclusion unit 213, and an additional selection unit 214.

  The intersection calculation unit 211 first obtains an encoded block from the preprocessing unit 100. When the motion vector calculation unit 113 divides the encoded block into prediction blocks, the intersection calculation unit 211 calculates an intersection between the division boundary line of the prediction block that is the current prediction target and the boundary line of the encoding block. Hereinafter, the current prediction target prediction block is referred to as a prediction target block. The intersection calculation unit 211 notifies the region dividing unit 212 of the calculated intersection.

  The area dividing unit 212 extends the dividing boundary line from the intersection point to the outer side of the encoded block with respect to the intersection point acquired from the intersection point calculating unit 211, and divides the intersection point into two regions using the extended dividing boundary line. These two regions are a region 1 including a prediction block and a region 2 including no prediction block. The area dividing unit 211 notifies the exclusion unit 213 of information on the divided areas 1 and 2. Below, the information of the divided | segmented area | regions 1 and 2 is called area | region information.

  The excluding unit 213 first acquires region information from the region dividing unit 212 and also acquires adjacent blocks adjacent to the prediction target block from the decoded image storage unit 110. The excluding unit 213 excludes adjacent blocks included in the area 2 from the peripheral candidate blocks based on the acquired area information. In other words, the excluding unit 213 also functions as a candidate selecting unit that selects adjacent blocks included in the region 1 as peripheral candidate blocks. The exclusion unit 213 outputs the selected peripheral candidate block to the inter prediction unit 112.

  Further, the excluding unit 213 may exclude an adjacent block in which a part of the block or the center of gravity of the block is included in the area not including the prediction target block (second area) from the peripheral candidate blocks. At this time, the exclusion unit 213 calculates the centroid of the adjacent blocks included in the region 2 and determines whether the centroid is included in the region 2.

  The inter prediction unit 112 determines the predicted motion vector as described above using the motion vector of the peripheral candidate block selected by the selection unit 201, and performs motion compensation using the predicted motion vector. The inter prediction unit 112 determines the coding performance based on the shapes of various prediction target blocks, and determines the shape of the prediction block having the best coding performance.

  Also, by excluding adjacent blocks by the exclusion unit 213, the neighboring candidate blocks may be equal to or less than a predetermined value. The predetermined value is any one of 0 to 4, for example. When the peripheral candidate blocks are equal to or smaller than the predetermined value, the peripheral candidate blocks are reduced, and therefore, there is a high possibility that appropriate motion vector prediction cannot be performed.

  Therefore, the additional selection unit 214 additionally selects an adjacent block that is a peripheral candidate block according to a predetermined criterion from the adjacent blocks once excluded. For example, the additional selection unit 214 selects adjacent blocks from the adjacent blocks included in the region 2 in the order closer to the prediction target block until a predetermined number is reached. The predetermined number may be the same number as the predetermined value, or may be a number greater than the predetermined value.

  Here, when the division is expressed in a hierarchy, the division from the encoded block to the prediction block is expressed as the division into the minimum hierarchy. For example, the division boundary line of the minimum hierarchy is a boundary line for dividing the prediction block, and the division boundary lines of the other hierarchy are boundary lines for dividing the encoded block.

  At this time, the intersection calculation unit 211 and the region division unit 212 can also be described as follows. The intersection calculation unit 211 calculates the intersection of the division boundary line at the minimum hierarchy and the division boundary line at another hierarchy.

  The area dividing unit 212 is an area composed of a half line drawn parallel to the division boundary line of the minimum hierarchy from the intersection calculated by the intersection calculation unit 211 and drawn outward from the coding block, and the division boundary line of the minimum hierarchy. Is divided into two.

  As a result, the selection unit 201 reduces the amount of calculation by narrowing down the peripheral candidate blocks based on the shape and position of the prediction target block, and sets adjacent blocks having movement similar to the prediction target block as the peripheral candidate blocks. By doing so, the prediction accuracy can be maintained.

  This is based on the idea that the fact that the coding block is divided into prediction blocks means that the prediction target block and other prediction blocks basically have different motions, so that the division is performed. That is, the motion of a coded block adjacent to another prediction block different from the prediction target block is likely to have a motion different from that of the prediction target block.

  Therefore, the region is divided into two using the division boundary line for dividing the prediction block, and the adjacent block included in the region including the prediction target block is estimated to have the same motion as the prediction target block. On the other hand, it is estimated that an adjacent block included in a region not including the prediction target block has a motion different from that of the prediction target block.

<Example of neighboring candidate blocks>
Next, an example of the peripheral candidate block selected according to the shape and position of the prediction target block will be described. FIG. 5 is a diagram for describing a first example of the peripheral candidate blocks. In the example illustrated in FIG. 5, a block b200 surrounded by a bold line indicates an encoded block, and a block b201 of satin 1 (the density of dots is lower than the satin 2) indicates a prediction target block.

  The broken line shown in FIG. 5 shows the boundary line for dividing | segmenting into a prediction block, and the thick broken line shows the division | segmentation boundary line of a prediction object block. Also, the intersections c101 and c102 shown in FIG. 5 indicate the intersections between the division boundary line (thick broken line) of the prediction target block and the boundary line (thick line) of the encoded block.

  Dotted lines d101 and d102 shown in FIG. 5 indicate lines obtained by extending the dividing boundary lines from the intersections c101 and c102 to the outside of the encoding block b200, respectively. The area is divided into two regions using an extended division boundary line (a thick broken line and a thick one-dot chain line).

  At this time, a region including the prediction target block b201 is referred to as a region 1, and a region not including the prediction target block b201 is referred to as a region 2.

  The excluding unit 213 excludes adjacent blocks b221 and b222 represented by diagonal lines included in the region 2 from the peripheral candidate blocks. The exclusion unit 213 also functions as a candidate selection unit that selects the adjacent blocks b211 to b213 represented in white included in the region 1 as peripheral candidate blocks.

  Here, since raster scanning is assumed, blocks b231 to 233 in satin 2 (the density of points is higher than that in satin 1) are not encoded, and therefore cannot be referenced in the prediction target block b201. It is.

  FIG. 6 is a diagram for describing a second example of the peripheral candidate blocks. In the example shown in FIG. 6, the shape of the prediction block is different from the shape of the prediction block shown in FIG. In addition, FIG. 6 illustrates which adjacent blocks are excluded at the position of the prediction target block. The definitions of the thick broken line, the thick dashed line, the thick line, the satin 1 block, the satin 2 block, the hatched block, and the white block shown in FIG. 6 are the same as the definitions shown in FIG.

  FIG. 6A shows an example of neighboring candidate blocks when divided into vertically predicted blocks and the left side is a prediction target block. Since the prediction block shown in FIG. 6A is divided in the vertical direction, it is estimated that the prediction blocks adjacent to the left and right move differently. Therefore, the prediction block adjacent to the right of the prediction target block and the adjacent block adjacent to the prediction block in the vertical direction are excluded from the peripheral candidate blocks.

  FIG. 6B shows an example of a peripheral candidate block when divided into vertically long prediction blocks and the right side is a prediction target block. Since the prediction block shown in FIG. 6 (B) is divided in the vertical direction, it is estimated that the prediction blocks adjacent to the left and right move differently. Therefore, the prediction block adjacent to the left of the prediction target block and the adjacent block adjacent to the prediction block in the vertical direction are excluded from the peripheral candidate blocks.

  FIG. 6C illustrates an example of a peripheral candidate block when divided into horizontally long prediction blocks and the upper side is a prediction target block. Since the prediction block shown in FIG. 6C is divided in the horizontal direction, it is estimated that the prediction blocks adjacent to each other vertically move differently. For this reason, the prediction block adjacent to the prediction target block and the adjacent block adjacent to the prediction block in the horizontal direction are excluded from the peripheral candidate blocks.

  FIG. 6D illustrates an example of a peripheral candidate block when divided into horizontally long prediction blocks and the lower side is a prediction target block. Since the prediction block shown in FIG. 6D is divided in the horizontal direction, it is estimated that the prediction blocks adjacent to each other vertically move differently. Therefore, the prediction block adjacent to the prediction target block and the adjacent block adjacent to the prediction block in the horizontal direction are excluded from the peripheral candidate blocks.

  FIG. 7 is a diagram for explaining a third example of the peripheral candidate blocks. In the example illustrated in FIG. 7, the shape of the prediction block is different from the shape of the prediction block illustrated in FIGS. 5 and 6. Note that the definitions of the thick broken line, the thick dashed line, the thick line, the satin 1 block, the satin 2 block, the shaded block, and the white block shown in FIG. 7 are the same as the definitions shown in FIG.

  Moreover, although the prediction block shown in FIG. 7 is not rectangular, a trapezoid as shown in FIG. 7 or other arbitrary shapes may be used as the prediction block. Therefore, in the case of a prediction block having an arbitrary shape as shown in FIG. 7, which neighboring block is excluded at the position of the prediction target block will be described with reference to FIG.

  FIG. 7A shows an example of the peripheral candidate block when the encoded block is divided so that one of the prediction blocks becomes a trapezoid and the left side is a prediction target block. The prediction target block shown in FIG. 7A is divided into irregular shapes, but it is estimated that the blocks included in the region 2 that does not include the prediction target block move differently according to the original criterion. Therefore, the adjacent block included in the region 2 is excluded from the peripheral candidate blocks.

  FIG. 7B shows the same prediction block division as that in FIG. 7A, and shows an example of a peripheral candidate block when the right region is a prediction target block. The prediction target block illustrated in FIG. 7B is estimated to move differently from the blocks included in the region 2. Therefore, the adjacent block included in the region 2 is excluded from the peripheral candidate blocks.

  Further, in the example illustrated in FIG. 7, the division boundary line of the lowest hierarchy (division boundary line of the prediction block) does not intersect perpendicularly with the division boundary line of the other hierarchy (the boundary line of the encoded block). The case where the dividing boundary line of the minimum hierarchy and the dividing boundary line of the other hierarchy intersect perpendicularly is the same as the above example.

  The direction estimation of the half line from the intersection when the division boundary line of the minimum hierarchy and the division boundary line of the other hierarchy do not intersect perpendicularly will be described with reference to FIG. FIG. 8 is a diagram for explaining the estimated direction of the half line. A region ar301 illustrated in FIG. 8 indicates a region used for estimating the direction of the half line. An intersection point c301 illustrated in FIG. 8 indicates an intersection point where the division boundary line of the minimum hierarchy and the division boundary lines of the other hierarchy do not intersect perpendicularly. A thick alternate long and short dash line shown in FIG. 8 indicates a half line from the intersection.

  At this time, the region dividing unit 212 estimates the direction of the half line d301 extending from the intersection c301 based on the number of pixels of the prediction target block in the constant pixel region ar301 near the intersection c301. For example, the area dividing unit 212 obtains a ratio between the number of pixels of the prediction target block and the number of pixels of the prediction block other than the prediction target block included in the area ar301, and based on this ratio, a half line from the intersection The inclination of d301 is determined.

  The area dividing unit 212 divides the image into two areas by a dividing boundary line of the minimum hierarchy or an extended dividing boundary line and a half line. Subsequent processing is the same as that described above.

  As described with reference to FIGS. 6 to 8, the peripheral candidate blocks estimated to have the same motion as the prediction target block can be selected depending on the shape and position of the prediction target block.

<Additional selection>
Next, additional selection of peripheral candidate blocks will be described with reference to the drawings. FIG. 9 is a diagram for explaining additional selection of neighboring candidate blocks. Blocks b421 and b422 illustrated in FIG. 9 are blocks excluded from the peripheral candidate blocks by the exclusion unit 213.

  Here, it is assumed that the additional selection unit 214 determines that the number of selected peripheral candidate blocks is equal to or less than a predetermined value. At this time, the additional selection unit 214 additionally selects the excluded adjacent blocks as the peripheral candidate blocks in the order closer to the prediction target block b401.

  For example, the additional selection unit 214 additionally selects peripheral candidate blocks in the order of adjacent blocks b411 and b412. In addition, although mentioned later, in the prediction block in Example 1, the motion vector prediction process and motion compensation of each prediction block after a division | segmentation can be performed in parallel.

  Thus, when processing of each prediction block is performed in parallel, the block b411 has no motion vector when processing the prediction target block b401. In the first embodiment, it is assumed that encoding is performed in raster scan order.

  Therefore, the additional selection unit 214 in this case adds only the encoded block b412 to the peripheral candidate blocks.

<Parallel processing>
Next, it will be described that the processing of each prediction block can be performed in parallel. FIG. 10 is a diagram for explaining parallel processing of prediction blocks. In the example shown in FIG. 10, the encoding block b500 is divided into vertically predicted blocks. In addition, since it is assumed that encoding is performed in the raster scan order, in the general prior art, the prediction target block b501 shown in FIG. 10A, and then the prediction target block b601 shown in FIG. 10B. Are encoded in this order.

  In the prior art, when processing the prediction target block b601, the motion vector of the adjacent block b621 is required, and therefore the prediction target block b501 needs to be encoded first.

  On the other hand, according to the first embodiment, as illustrated in FIG. 10B, when the prediction target block b601 is processed, the adjacent block b621 is excluded, so that the motion vector of the adjacent block b621 is not required.

  The adjacent block used in the prediction target block b601 shown in FIG. 10B has already been encoded when the prediction target block b501 shown in FIG. 10A is processed. Therefore, the prediction target block b501 shown in FIG. 10A and the prediction target block b601 shown in FIG. 10B can be processed in parallel.

  If the method for selecting neighboring candidate blocks in the first embodiment is used, the same can be said for the division of prediction blocks having various shapes regardless of the shape of the prediction block shown in FIG. Therefore, according to Example 1, since the process of each prediction block can be parallelized, it can speed up rather than a prior art.

<Operation>
Next, the operation of the image encoding device 10 in the first embodiment will be described. FIG. 11 is a flowchart illustrating an example of motion vector prediction processing. In step S101 illustrated in FIG. 11, the selection unit 201 selects an adjacent block included in a region that does not include the prediction target block from the encoded adjacent blocks adjacent to the prediction block, based on the division shape and position of the prediction target block. Neighboring candidate blocks are selected except for.

  In step S102, the inter prediction unit 112 has the best prediction among the motion compensation based on the motion vector obtained by the motion vector calculation based on the conventional block matching and the plurality of peripheral candidate blocks selected by the selection unit 201. Coding performance is compared for motion compensation using motion vectors calculated by combinations of neighboring candidate blocks. As a result of comparison, if the motion compensation encoding performance by block matching is high, the motion vector coefficient is output from the motion vector calculation unit 113 to the entropy encoding unit 104, and the motion compensation encoding performance by motion vector prediction is high. In this case, the index of the best neighboring candidate block is output from the motion vector calculation unit 113 to the entropy encoding unit 104 and included in the stream.

  Next, the peripheral candidate block selection process will be described in detail. FIG. 12 is a flowchart illustrating an example of a peripheral candidate block selection process according to the first embodiment. In step S201 illustrated in FIG. 12, the selection unit 201 inputs an encoded block.

  In step S202, the selection unit 201 searches for a referenceable block from the decoded image storage unit 110, and acquires information regarding the referenceable block. The information regarding the block includes block position information and motion vector information.

  In step S <b> 203, the intersection calculation unit 211 calculates the intersection between the division boundary line of the minimum hierarchy (division boundary line of the prediction block) and the division boundary line of another hierarchy (the boundary line of the encoded block).

  In step S204, the area dividing unit 212 determines an excluded area (area 2) to be excluded from the peripheral candidate blocks. The area dividing unit 212 extends, for example, the dividing boundary line from the intersection calculated by the intersection calculating unit 211 in the outer direction of the encoded block, and divides the divided area into two regions using the extended dividing boundary line. At this time, the region dividing unit 212 determines a region that does not include the prediction target block as an excluded region.

  In step S205, the exclusion unit 213 excludes adjacent blocks included in the exclusion region from the peripheral candidate blocks.

  In step S206, the exclusion unit 213 calculates the position of the center of gravity of the adjacent block. In step S207, the exclusion unit 213 determines whether or not the centroid of the adjacent block is included in the exclusion region. If the center of gravity is included in the excluded area (step S207—YES), the process proceeds to step S208. If the center of gravity is not included in the excluded area (step S207—NO), the process proceeds to step S209.

  However, if the center of gravity of the adjacent block is on the extended division boundary line, it may be set in advance whether or not to exclude.

  In step S208, the exclusion unit 213 excludes adjacent blocks whose center of gravity is included in the exclusion region from the peripheral candidate blocks.

  In step S209, the selection unit 201 selects adjacent blocks that remain without being excluded as peripheral candidate blocks, and outputs the selected peripheral candidate blocks to the inter prediction unit 112.

  In addition, before the process of step S209, the additional selection part 214 may perform the additional selection process of a periphery candidate block. Moreover, the process of step S206-S208 is not necessarily a required process.

  As described above, according to the first embodiment, when selecting neighboring candidate blocks by motion vector prediction, it is possible to reduce the calculation amount while preventing the prediction accuracy from being lowered. Moreover, since the process of each prediction block in Example 1 can be performed in parallel, it leads to high speed. Further, according to the first embodiment, the prediction block is not limited to a rectangular block.

[Example 2]
FIG. 13 is a block diagram illustrating an example of the configuration of the image processing apparatus 30 according to the second embodiment. An image processing apparatus 30 illustrated in FIG. 13 is an example of an apparatus in which the image encoding process described in the first embodiment is implemented by software.

  As illustrated in FIG. 13, the image processing apparatus 30 includes a control unit 301, a main storage unit 302, an auxiliary storage unit 303, a drive device 304, a network I / F unit 306, an input unit 307, and a display unit 308. These components are connected to each other via a bus so as to be able to transmit and receive data.

  The control unit 301 is a CPU that controls each device, calculates data, and processes in a computer. The control unit 301 is an arithmetic device that executes an image encoding processing program stored in the main storage unit 302 or the auxiliary storage unit 303. The control unit 301 receives data from the input unit 307 and the storage device, calculates and processes the data, and outputs the data to the display unit 308 and the storage device.

  The control unit 301 can realize the processing described in each embodiment by executing a program for image encoding processing.

  The main storage unit 302 is a ROM (Read Only Memory), a RAM (Random Access Memory), or the like. The main storage unit 302 is a storage device that stores or temporarily stores programs and data such as OS (Operating System) and application software that are basic software executed by the control unit 301.

  The auxiliary storage unit 303 is an HDD (Hard Disk Drive) or the like, and is a storage device that stores data related to application software and the like.

  The drive device 304 reads the program from the recording medium 305, for example, a flexible disk, and installs it in the storage unit.

  A predetermined program is stored in the recording medium 305, and the program stored in the recording medium 305 is installed in the image processing apparatus 30 via the drive device 304. The installed predetermined program can be executed by the image processing apparatus 30.

  The network I / F unit 306 is a peripheral having a communication function connected via a network such as a LAN (Local Area Network) or a WAN (Wide Area Network) constructed by a data transmission path such as a wired and / or wireless line. This is an interface between the device and the image processing apparatus 30.

  The input unit 307 includes a keyboard having cursor keys, numeric input, various function keys, and the like, a mouse and a slice pad for selecting keys on the display screen of the display unit 308, and the like. The input unit 307 is a user interface for a user to give an operation instruction to the control unit 301 or input data.

  The display unit 308 is configured by a CRT (Cathode Ray Tube), an LCD (Liquid Crystal Display), or the like, and performs display according to display data input from the control unit 301.

  The decoded image storage unit 110 illustrated in FIG. 3 is realized by, for example, the main storage unit 302 or the auxiliary storage unit 303. The configuration other than the decoded image storage unit 110 illustrated in FIG. 3 includes, for example, a control unit 301 and a work memory. This can be realized by the main storage unit 302.

  The program executed by the image processing apparatus 30 has a module configuration including each unit described in the first embodiment. As actual hardware, when the control unit 301 reads a program from the auxiliary storage unit 303 and executes it, one or more of the above-described units are loaded onto the main storage unit 302, and one or more of the units are main. It is generated on the storage unit 302.

  As described above, the image encoding processing described in the first embodiment may be realized as a program for causing a computer to execute. The processing described in each embodiment can be realized by installing this program from a server or the like and causing the computer to execute the program.

  It is also possible to record the program on the recording medium 305 and cause the computer or portable terminal to read the recording medium 305 on which the program is recorded to realize the above-described image encoding process. The recording medium 305 is a recording medium that records information optically, electrically, or magnetically, such as a CD-ROM, a flexible disk, or a magneto-optical disk, and information is electrically stored such as a ROM or flash memory. Various types of recording media such as a semiconductor memory for recording can be used.

  Moreover, various integrated circuits and electronic circuits can be employed for each unit described in the first embodiment. In addition, a part of each part described in the first embodiment can be another integrated circuit or an electronic circuit. For example, examples of the integrated circuit include ASIC (Application Specific Integrated Circuit) and FPGA (Field Programmable Gate Array). Examples of the electronic circuit include a central processing unit (CPU) and a micro processing unit (MPU).

  Each embodiment has been described in detail above. However, the present invention is not limited to the specific embodiment, and various modifications and changes other than the above-described modification are possible within the scope described in the claims. .

DESCRIPTION OF SYMBOLS 10 Image coding apparatus 30 Image processing apparatus 101 Predictive image generation part 102 Orthogonal transformation part 103 Quantization part 104 Entropy encoding part 105 Inverse quantization part 106 Inverse orthogonal transformation part 107 Decoded image generation part 108 Deblocking filter part 109 Loop filter Unit 110 decoded image storage unit 111 intra prediction unit 112 inter prediction unit 113 motion vector calculation unit 114 encoding control and header generation unit 115 prediction image selection unit 201 selection unit 211 intersection calculation unit 212 region division unit 213 exclusion unit 214 additional selection unit 301 Control Unit 302 Main Storage Unit 303 Auxiliary Storage Unit 304 Drive Device

Claims (5)

An image coding apparatus that divides an image into coding blocks, divides the coding block into prediction blocks that are prediction targets of motion vector prediction, and performs coding of the image,
A selection unit that selects a coded block adjacent to the prediction block based on a division shape and a position of the prediction block in the coding block;
An inter prediction unit that performs motion compensation of the prediction block using a motion vector of the encoded block selected by the selection unit;
An image encoding device comprising: The selection unit includes:
When the coding block is divided into the prediction blocks, an intersection calculation unit that calculates an intersection between a division boundary line of the prediction block that is a current prediction target and a boundary line of the coding block;
An area dividing unit that extends the dividing boundary line from the intersection point toward the outside of the coding block, and divides the dividing boundary line into two areas using the extended dividing boundary line;
An excluding unit that excludes the encoded block included in a region that does not include the current prediction target prediction block from the two regions, from selection candidates;
The image encoding device according to claim 1, further comprising: The exclusion part is
The image encoding apparatus according to claim 2, wherein an encoded block in which a part of the block or the center of gravity of the block is included in an area not including the current prediction target prediction block is also excluded from the selection candidates. The selection unit includes:
When the number of the selected encoded blocks is equal to or less than a predetermined value, the encoded blocks included in the area not including the current prediction target prediction block are close to the current prediction target prediction block. The image encoding device according to claim 2 or 3, wherein the image encoding device sequentially selects until the number of the encoded blocks reaches a predetermined number. A computer that divides an image into coding blocks, divides the coding block into prediction blocks that are prediction targets for motion vector prediction, and encodes the image,
A selection step of selecting an encoded block adjacent to the prediction block based on a division shape and a position of the prediction block in the encoding block;
An inter prediction step of performing motion compensation of the prediction block using a motion vector of the selected encoded block;
A program for running

Images