JP2016187171A - Video encoding device, video encoding method, and video encoding program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To settle the mismatch of vector copy relationship occurring in the subblock of peripheral copy mode, when changing the coding mode of some subblock from inter coding mode to intra coding mode, after determination of the coding mode.SOLUTION: A video encoding device includes a determination unit and a change unit. The determination unit determines the coding mode applied, respectively, to a plurality of blocks in an image, and determines a prediction mode to be applied. When the coding mode of the first block is changed to the intra-coding mode, the prediction mode of the second block adjacent to the first block has been determined to a copy mode for copying the motion vector of an adjacent block, after the coding mode of the first block, out of a plurality blocks, is determined to the inter coding mode by the determination unit, the change unit changes the prediction mode of the second block to a prediction vector mode for encoding the difference of the motion vector and prediction vector.SELECTED DRAWING: Figure 5

Description

  The present invention relates to a moving image encoding apparatus, a moving image encoding method, and a moving image encoding program.

  Since image data, particularly moving image data, often has a large amount of data, for example, when data is transmitted from a moving image encoding device to a moving image decoding device or stored in a storage device, a data string Is encoded and converted into another data string. Thereby, for example, the data amount is compressed and transmitted from the moving image encoding device to the moving image decoding device, thereby enabling smooth browsing on the receiving side moving image decoding device.

  As a technique for encoding a moving image, for example, H.264. Technologies such as H.264 and High Efficiency Video Coding (HEVC, H.265) are known.

  When encoding a moving image, the image may be divided into a plurality of blocks, and encoding may be performed in units of blocks. In this case, the block may be further divided into a plurality of sub blocks, and the encoding mode may be determined in units of sub blocks. For example, in HEVC, an image is divided into blocks called Coding Tree Units (CTUs), and the CTU is further divided into a plurality of sub-blocks called Coding Units (CUs). Then, the encoding mode is selected on a CU basis. Note that the encoding mode may include, for example, an inter encoding mode and an intra encoding mode.

  The inter coding mode is a coding mode in which a coding target image included in a moving image is coded using information of a coded image, and the intra coding mode is information that the coding target image has. This is an encoding mode for encoding the target image. Note that the encoding target image may be referred to as a frame or a picture in the following description.

  In many cases, moving image data has a high degree of similarity between picture data at a certain timing and picture data at the next timing. For this reason, in the inter coding mode, coding is performed using the property of high correlation in the time direction of the moving image data. First, an image is divided into a plurality of blocks. Then, for each block, the moving image encoding apparatus selects a region similar to the target block to be encoded from the decoded image of the encoded frame, and specifies a motion vector indicating the similar region. In addition, the moving image encoding apparatus calculates a difference between the similar region and a block to be encoded. Then, the moving image encoding apparatus realizes a high compression rate by encoding the information indicating the identified motion vector and the calculated difference information.

  Here, the inter coding mode may include two prediction modes, for example, a copy mode and a prediction vector mode. The copy mode may be a mode for copying the motion vector of the encoded peripheral block. For example, HEVC defines a merge mode as a copy mode. In the merge mode, a plurality of motion vector candidates are derived using motion vectors of encoded adjacent blocks, and a merge candidate list is generated. Then, a suitable motion vector candidate is selected from the merge candidate list, and an index designating the selected motion vector candidate is encoded and transmitted. The prediction vector mode may be a mode in which a difference vector between a motion vector of a block to be processed and a prediction vector derived from a motion vector of an encoded adjacent block is encoded. For example, in HEVC, Advanced Motion Vector Prediction (AMVP) is defined as a prediction vector mode. Also, for example, H. Also in H.264, a copy mode for copying a motion vector of an encoded adjacent block and a prediction vector mode for encoding a difference vector from the motion vector using the encoded adjacent block as a prediction vector are defined.

  In this regard, there is known a technique for detecting a motion vector with high accuracy when there is a block encoded in the intra mode among the peripheral blocks of a block that is a target of motion vector detection in motion detection during moving image encoding. It has been. (For example, Patent Document 1) When a virtual motion vector is allocated to a block encoded in the intra prediction mode and a motion vector of a block encoded in the inter prediction mode is encoded after the intra block, the block is allocated to the intra block. A technique using a virtual motion vector is known. (For example, patent document 2) Moreover, the technique relevant to patent documents 3-7 is described.

JP 2006-20095 A Special table 2010-517405 gazette JP 2014-103429 A JP 2014-33460 A JP 2012-191247 A JP 2012-191491 A JP 2012-191397 A

  Here, for example, in order to realize real-time processing, the moving image encoding apparatus may collectively determine the encoding mode and the prediction mode for a plurality of sub-blocks in a certain block. Then, after the determination of the encoding mode, the moving image encoding apparatus may change, for example, the sub-block determined to be the inter encoding mode to the intra encoding mode. In such a case, for example, the coding mode of another subblock adjacent to the subblock changed from the inter coding mode to the intra coding mode is inter coding, and the prediction mode is copy mode. Suppose. In this case, there is a possibility that the sub-block to be referred to for deriving the motion vector to be copied in the copy mode includes a sub-block whose coding mode is changed from the inter coding mode to the intra coding mode. There is. However, when the mode is changed to the intra coding mode, the referenced sub-block does not have a motion vector. Therefore, the motion vector of the sub block changed to the intra coding mode is not referred to. As a result, the copy relationship of motion vectors in the copy mode is broken, and an incorrect motion vector may be selected in another adjacent sub-block. Furthermore, when the prediction mode of a sub-block adjacent to another sub-block from which this erroneous vector is selected is also a copy mode, the motion vector is derived using the chained erroneous motion vector. . As a result, the error further propagates to surrounding sub-blocks, and the video decoding device that is the transmission destination of the video data may not decode the stream correctly, which may cause problems such as image quality degradation. An object according to one aspect is that a vector generated in a sub-block of a peripheral copy mode when a coding mode of a certain sub-block is changed from an inter-coding mode to an intra-coding mode after the coding mode is determined. It is to resolve inconsistencies in copy relationships.

  The moving picture coding apparatus according to one aspect of the present invention includes a determination unit and a change unit. The determination unit determines an encoding mode to be applied to each of the plurality of blocks in the image and determines a prediction mode to be applied to each of the plurality of blocks. The changing unit changes the coding mode of the first block to the intra coding mode after the coding mode of the first block among the plurality of blocks is determined to the inter coding mode by the determining unit, When the prediction mode of the second block adjacent to the first block among the blocks is determined to be the copy mode for copying the motion vector of the adjacent block, the prediction mode of the second block is changed from the copy mode to the motion vector. And a prediction vector mode for encoding a difference vector between the prediction vector and the prediction vector.

  According to one aspect, when the coding mode of a certain sub-block is changed from the inter-coding mode to the intra-coding mode after the coding mode is determined, the vector copy that occurs in the sub-blocks in the neighboring copy mode Relationship inconsistencies can be resolved.

It is a figure which illustrates the structure of the block in HEVC. It is a figure which illustrates about the mismatching which may arise when changing the subblock determined to inter coding mode to intra coding mode. It is a figure which illustrates the functional block structure of the moving image encoder which concerns on 1st Embodiment. It is a figure which illustrates the flow of the encoding process of the moving image which concerns on 1st Embodiment. It is a figure which illustrates the operation | movement flow of the encoding process of the moving image which concerns on 1st Embodiment. It is a figure which illustrates the change to the IPCM of the encoding mode of a subblock, and the change to the prediction vector mode of a surrounding subblock. It is a figure which illustrates the subblock referred in copy mode. It is a figure which illustrates the operation | movement flow of the encoding process of the moving image which concerns on 2nd Embodiment. It is a figure which illustrates the hardware constitutions for realizing the moving picture coding device concerning one embodiment.

  Hereinafter, some embodiments of the present invention will be described in detail with reference to the drawings. In addition, the same code | symbol is attached | subjected to the corresponding element in several drawing. In the following, a case where HEVC is used as a moving image encoding method will be described as an example. However, the moving picture encoding method that can be used in the embodiment is not limited to this. For example, H.M. In other moving image coding systems such as H.264 and MPEG-2, a coding mode of a plurality of sub-blocks (for example, a macro block) in a certain block is determined in a lump, and then the coding mode is changed. If so, embodiments can be applied as well.

  FIG. 1 is a diagram illustrating a block configuration in a moving image encoding scheme: HEVC. In HEVC, as shown in FIG. 1, a picture is divided into blocks called CTUs. The CTU is encoded in the raster scan order. The CTU size is constant in sequence units, and may be set to any size of 16 × 16 pixels, 32 × 32 pixels, or 64 × 64 pixels, for example. The CTU is further divided into CUs with a quadtree structure, and the divided CUs are encoded in the Z-scan order. The size of the CU is variable. For example, a size of 8 × 8 pixels to 64 × 64 pixels can be selected. The CU is a unit for selecting a coding mode, and the coding mode may include, for example, an intra coding mode for performing intra prediction coding and an inter coding mode for performing inter prediction coding.

  The CU is individually processed for each Prediction Unit (PU) and Transform Unit (TU). The PU is a unit for determining the prediction mode in each coding mode. For example, the intra coding mode is a unit for determining the intra prediction mode, and the inter coding mode is a unit for performing motion compensation. For example, if the PU size is inter prediction, the PU partition shape is 2N × 2N, N × N, 2N × N, N × 2N, 2N × nU, 2N × nD, nR × 2N, nL × 2N. You can choose from different PU partition types. The PU division shape names nU, nD, nR, and nL indicate that the smaller PU is located at the top, bottom, right, and left in the CU, respectively. On the other hand, TU is a unit of orthogonal transform, and TU size can select from 4 × 4 pixels to 32 × 32 pixels. In the TU division, the CU is divided in a quadtree structure and processed in the Z scan order. In the following description, the CTU may be referred to as a coding block. In addition, blocks below the CTU (for example, CU, PU, and TU) may be referred to as sub-blocks.

  In HEVC, the motion vector can be obtained in units of PUs, and the prediction mode can be determined in units of PUs. In HEVC, as described above, two prediction modes of a prediction vector mode called AMVP and a copy mode called a merge mode are defined as motion vector encoding methods. AMVP may be a mode specified by syntax: merge_flag = 0, for example. In addition, the merge mode may be a mode specified by the syntax: cu_skip_flag = 1 or merge_flag = 1, for example.

  Here, for example, in order to realize real-time processing, the moving image encoding apparatus collectively sets the encoding mode and the prediction mode for each of a plurality of sub-blocks (for example, CU, PU, etc.) in one CTU. May be determined. Further, since the subsequent encoding process is performed according to the standard, for example, as shown in FIG. 1, the CUs in the CTU are encoded in the Z scan order. However, during this encoding process, the encoding target CU that has been determined to be the inter encoding mode may be changed to Intra Pulse Code Modulation (IPCM) or another intra encoding mode. Note that the IPCM may be a kind of prediction mode in the intra coding mode in which PCM data of a sub-block (for example, CU) is transmitted.

  Examples of changing the encoding target CU that has been determined to be the inter encoding mode to the IPCM or another intra encoding mode during the encoding process are as follows. For example, in HEVC, a limit value is provided for the code amount per CTU based on the code amount of IPCM. On the other hand, the encoding mode (for example, IPCM) can be selected in units of CUs. Therefore, a threshold value normalized for each CU is determined based on the limit value of the code amount set per CTU. Subsequently, the threshold value determined for the CU is compared with the code amount (cost) when encoding is performed in the encoding mode determined for the processing target sub-block. When the code amount exceeds the determined threshold value, the encoding mode of the sub-block to be processed is replaced with IPCM, so that the encoding cost is suppressed in each CU and the limit value in CTU units is satisfied. Is done. For example, in this case, the encoding target sub-block that has been determined as the inter encoding mode may be changed to the IPCM. When the encoding target CU that has been determined to be the inter encoding mode is changed to the IPCM or another intra encoding mode, the sub-block encoded thereafter is caused by the change of the encoding mode. Inconsistencies may occur.

  FIG. 2 is a diagram illustrating an inconsistency that may occur when a sub-block to be encoded that has been determined to be in the inter encoding mode is changed to an IPCM or another intra encoding mode. FIG. 2A is a diagram exemplifying a state in which the moving image encoding apparatus collectively determines the encoding mode and the prediction mode for each of a plurality of sub-blocks in the CTU. In FIG. 2A, it is assumed that the coding mode of a plurality of sub-blocks (for example, CU) included in the CTU is determined as the inter coding mode or the intra coding mode. In FIG. 2A, for each PU included in the CU, the prediction mode is determined to be a merge mode that is a copy mode in HEVC or AMVP that is a prediction vector mode in HEVC. In this state, it is assumed that the moving image encoding apparatus encodes the CUs in the CTU in the Z scan order according to the determined encoding mode and prediction mode. Here, the sub-block indicated by the character “target” in FIG. 2A represents the current sub-block to be encoded, and the encoding mode is the inter encoding mode (for example, merge). Mode or AMVP). In this case, for example, the amount of code when encoding is performed in the encoding mode determined for the sub-block to be encoded exceeds a threshold normalized with respect to the CU. Assume that the encoding mode of the sub-block to be converted is changed to IPCM. In this case, as shown in FIG. 2B, inconsistency may occur in the neighboring sub-blocks (for example, PUs) having a dependency relationship with the coding mode of the changed sub-block.

  As shown in FIG. 2B, the sub-block (for example, CU) indicated by the characters “target” in FIG. 2A is changed to IPCM which is one of the prediction modes of the intra coding mode. To do. Also, it is assumed that the prediction mode of a subblock (for example, PU) adjacent to a subblock changed to IPCM is a merge mode that is a copy mode for copying a motion vector of an encoded adjacent subblock in HEVC. In this case, the motion vector prediction may change due to the change of the encoding mode. For example, in the merge mode in HEVC, a plurality of motion vector candidates are derived using motion vectors of encoded adjacent blocks, and a merge candidate list is generated. In the merge candidate list, a reference sub-block that is a copy source of a motion vector is represented by a syntax called merge_idx, and merge_idx is related to a surrounding sub-block by a method defined in the HEVEC standard. . In the merge mode, a suitable motion vector candidate is selected from the merge candidate list, and merge_idx designating the selected motion vector candidate is encoded and transmitted. In addition, the moving image decoding device on the receiving side acquires a motion vector based on merge_idx and decodes the moving image.

  However, when the peripheral subblock (for example, CU) of the merge mode subblock (for example, PU) is changed to IPCM, the relationship between the merge candidate list and the peripheral subblock changes, that is, indicated by merge_idx. The motion vector changes. Therefore, as shown in FIG. 2B, an incorrect motion vector is selected in a merge mode sub-block (for example, PU) adjacent to the sub-block (for example, CU) changed to IPCM. As a result, image quality degradation or the like may occur when decoding is performed by the video decoding device on the receiving side. Further, if the sub-blocks around the sub-block for which the wrong motion vector is selected are also in the merge mode, the wrong motion vector is further copied, and the error may be propagated.

  Further, for example, it may be possible to cope with this problem by recalculating the merge candidate list of all the sub-blocks in the range in which the error propagates, but this requires a calculation amount. Furthermore, even if recalculation is performed, if the sub-block indicated by merge_idx has been changed to IPCM, the original motion vector may not exist. Eventually, the motion search is performed again, ensuring real-time performance. There is a risk of being lost. For this reason, if a certain sub-block is changed from the inter coding mode to another coding mode after the coding modes of a plurality of sub-blocks in the block are determined, this may occur in the sub-blocks in the neighboring copy mode. A technique for resolving the alignment is desired.

  Therefore, for example, it is assumed that the sub-block to be encoded (for example, PU) is in the copy mode, and there are sub-blocks (for example, CU) that have been changed from the inter encoding mode to the intra encoding mode in the vicinity. . In this case, in the first embodiment, the moving image encoding apparatus changes, for example, the encoding mode of the sub-block to be encoded to the prediction vector mode. In the prediction vector mode, the motion vector is explicitly encoded. Therefore, even if the coding mode is changed to IPCM in a certain sub-block after determining the coding mode of a plurality of sub-blocks in the block, the moving picture coding apparatus solves inconsistency caused by the change. can do. Therefore, for example, a moving image can be appropriately decoded in the moving image decoding device on the receiving side. Hereinafter, the first embodiment will be described with reference to FIGS. 3 to 7.

<First Embodiment>
FIG. 3 is a diagram illustrating a functional block configuration of the video encoding device 30 according to the first embodiment. The moving image encoding device 30 includes, for example, a control unit 300 and a storage unit 310. The control unit 300 may include, for example, a determination unit 301 and a change unit 302. In addition, the control unit 300 includes a prediction error signal generation unit 401, an orthogonal transformation unit 402, a quantization unit 403, an entropy coding unit 404, an inverse quantization unit 405, an inverse orthogonal transformation unit 406, and a decoding shown in FIG. An image generation unit 407 and the like may be included. Furthermore, the control unit 300 may include, for example, an intra prediction image generation unit 409, an inter prediction image generation unit 410, a motion vector calculation unit 411, an encoding mode determination unit 412, a prediction image selection unit 413, and the like. The control unit 300 may include a copy mode determination unit 414, an inter encoding mode change unit 415, an adjacent subblock determination unit 416, an encoding mode change determination unit 418, and the like. In one embodiment, the control unit 300 of the video encoding device 30 may operate as these functional units by reading and executing a video encoding program using the storage unit 310, for example. Further, the determination unit 301 and the change unit 302 illustrated in FIG. 3 may correspond to, for example, some of the functional blocks illustrated in FIG. The storage unit 310 of the moving image encoding device 30 may store information such as a moving image encoding program and moving image data, for example. The storage unit 310 may function as, for example, a decoded image storage unit 408 and an intra block position storage unit 417 described later. Details of each functional unit included in the control unit 300 will be described later.

  FIG. 4 is a diagram illustrating a flow of a moving image encoding process executed by the functional block of the control unit 300 of the moving image encoding device 30 according to the first embodiment. In FIG. 4, the control unit 300 of the video encoding device 30 includes, for example, a prediction error signal generation unit 401, an orthogonal transform unit 402, a quantization unit 403, an entropy coding unit 404, an inverse quantization unit 405, and an inverse orthogonal transform. 406 and a decoded image generation unit 407. In addition, the control unit 300 of the video encoding device includes, for example, an intra prediction image generation unit 409, an inter prediction image generation unit 410, a motion vector calculation unit 411, a coding mode determination unit 412, and a prediction image selection unit 413. It is out. Furthermore, the control unit 300 of the moving image encoding apparatus 30 includes a copy mode determination unit 414, an inter encoding mode change unit 415, an adjacent sub-block determination unit 416, and an encoding mode change determination unit 418. The storage unit 310 may function as, for example, the decoded image storage unit 408 and the intra block position storage unit 417.

  The prediction error signal generation unit 401 divides a coding block to be coded included in a picture in moving image data into a plurality of sub-blocks according to a coding mode determined by a coding mode determination unit 412 described later. To do. The encoded block may be a CTU in HEVC, for example. Further, the sub-block may include, for example, CU, PU, and TU in HEVC. The prediction error signal generation unit 401 generates a prediction error signal using the processing target sub-block data obtained by the division and the prediction image sub-block data supplied from the prediction image selection unit 413. The prediction error signal generation unit 401 passes the generated prediction error signal to the orthogonal transformation unit 402.

  The orthogonal transform unit 402 performs orthogonal transform on the input prediction error signal. The orthogonal transform unit 402 supplies a signal separated into frequency components in the horizontal direction and the vertical direction by performing orthogonal transform on the prediction error signal to the quantization unit 403. The quantization unit 403 quantizes the output of the orthogonal transform unit 402. The quantization unit 403 reduces the code amount of the prediction error signal by encoding the signal by quantization, and supplies the signal to the copy mode determination unit 414 and the inverse quantization unit 405.

  The inverse quantization unit 405 dequantizes the output of the quantization unit 403 and then supplies the output to the inverse orthogonal transform unit 406. The inverse orthogonal transform unit 406 performs an inverse orthogonal transform process on the output of the inverse quantization unit 405 and then supplies the output to the decoded image generation unit 407. By performing decoding processing by the inverse quantization unit 405 and the inverse orthogonal transform unit 406, a signal comparable to the prediction error signal before encoding is obtained.

  For example, the decoded image generation unit 407 receives the sub-block data of the picture motion-compensated by the inter prediction image generation unit 410 and the prediction error signal decoded by the inverse quantization unit 405 and the inverse orthogonal transform unit 406. to add. As a result, the decoded image generation unit 407 reproduces the predicted sub-block data of the picture to be encoded and passes it to the decoded image storage unit 408.

  The decoded image storage unit 408 stores the passed predicted sub-block data as new reference picture data. The new reference picture data stored in the decoded image storage unit 408 may be used by, for example, the intra prediction image generation unit 409, the inter prediction image generation unit 410, and the motion vector calculation unit 411.

  The intra-predicted image generation unit 409 generates a predicted image from peripheral pixels that have already been encoded in the same picture. On the other hand, the inter predicted image generation unit 410 performs motion compensation on the reference picture data obtained from the decoded image storage unit 408 with the motion vector provided from the motion vector calculation unit 411, thereby subblocks of the motion-compensated reference picture. Generate data for

  The motion vector calculation unit 411 uses data of a coding block (for example, CTU in HEVC) in a picture to be coded and data of a coding block of an already coded reference picture obtained from the decoded image storage unit 408. To obtain a motion vector. The motion vector is, for example, a spatial shift in block units obtained by using a block matching technique (also referred to as motion vector search) that searches for a position most similar to a picture to be encoded from a reference picture in block units. Is a value indicating The motion vector calculation unit 411 passes the obtained motion vector to the inter predicted image generation unit 410.

  The coding mode determination unit 412 divides the coding block into a plurality of subblocks, and determines the coding mode of the subblock and the prediction mode belonging to the coding mode. The encoding mode may include, for example, an intra encoding mode and an inter encoding mode. Then, the sub-block data output from the intra predicted image generation unit 409 and the inter predicted image generation unit 410 according to the determined encoding mode and prediction mode are input to the predicted image selection unit 413. The predicted image selection unit 413 can select data of a sub-block of one of the predicted images. Data of the selected sub-block is supplied to the prediction error signal generation unit 401.

  Also, the encoding mode determination unit 412 determines the motion vector of the sub-block when determining the encoding mode of the sub-block. This determination may be executed, for example, by the motion vector calculation unit 411 performing a motion vector search as described above. Then, after the encoding mode determination unit 412 determines the encoding modes of all subblocks in the processing target encoding block, the following processing is executed for each subblock.

  First, an encoding mode determination unit includes an encoding mode determined for all sub-blocks included in an encoding target encoding block and information (for example, a prediction mode and a motion vector) according to the encoding mode. 412 is input to the copy mode determination unit 414. In addition, the copy mode determination unit 414 also receives the data of the processing target sub-block quantized by the quantization unit 403. Then, the copy mode determination unit 414 determines whether or not the sub-block to be processed is a copy mode in the inter coding mode based on the coding mode and prediction mode information input from the coding mode determination unit 412, for example. Determine whether. When the processing target sub block is in the copy mode, the copy mode determination unit 414 outputs information indicating the position of the processing target sub block to the adjacent sub block determination unit 416, for example. Also, the copy mode determination unit 414 outputs the data of the sub-block to be processed to the inter coding mode change unit 415.

  The adjacent sub-block determining unit 416 determines whether or not a sub-block whose encoding mode has been changed to IPCM is included in a sub-block adjacent to the copy-mode sub-block to which the position information is input. As will be described later, the intra block position storage unit 417 stores information on the position of the sub block changed from the inter coding mode to the IPCM. Whether the adjacent sub-block determining unit 416 includes the sub-block changed to the IPCM in the adjacent sub-block based on the position information of the sub-block changed to the IPCM stored in the intra block position storage unit 417. It may be determined whether or not. When the subblock changed to the IPCM is included in the subblock adjacent to the processing target subblock, the adjacent subblock determination unit 416 outputs information indicating that the condition is met to the inter coding mode change unit 415. .

  The inter coding mode changing unit 415 determines whether or not the sub-block to be processed is in the copy mode and the sub-block changed to the IPCM is included in the sub-block adjacent to the sub-block to be processed. . When the sub-block to be processed is in the copy mode and the sub-block changed to the IPCM is included in the sub-block adjacent to the sub-block to be processed, the sub-block to be processed is set to the prediction vector mode. Change to Note that the inter coding mode changing unit 415 may change the processing target sub-block to the prediction vector mode when information indicating that the condition is met is input from the adjacent sub-block determining unit 416, for example. .

  The motion vector used in the prediction vector mode to be changed includes a copy source in the copy mode of the sub-block to be processed before the encoding mode of the sub-block in the encoding block is changed by the encoding mode change determination unit 418 described later. May be set. As described above, the encoding mode determination unit 412 determines the encoding mode of each of the plurality of sub-blocks in the encoding block and information (for example, prediction mode, motion vector) according to the encoding mode. ing. Then, an encoding mode change determination unit 418, which will be described later, changes the encoding mode determined by the encoding mode determination unit 412. Therefore, the inter coding mode change unit 415 sets the copy source motion vector in the copy mode determined by the coding mode determination unit 412 for the processing target sub-block as the motion vector in the prediction vector mode to be changed. Good. In the following description, the copy-source motion vector in the copy mode that has been determined for the sub-block to be processed may be referred to as a copy vector.

  Further, this copy vector is a decoded image generated by the decoded image generation unit 407 in the process until the data of the processing target sub-block input to the prediction error signal generation unit 401 is input to the copy mode determination unit 414, for example. It may be stored in the storage unit 408. Then, the inter encoding mode changing unit 415 may acquire a copy vector in the copy mode determined by the encoding mode determining unit 412 for the processing target sub-block from the decoded image storage unit 408. Alternatively, in another embodiment, the encoding mode determination unit 412 determines when determining the encoding mode of the processing target sub-block and information (for example, prediction mode, motion vector) according to the encoding mode. The information may be stored in the storage unit 310, for example. Then, the inter coding mode changing unit 415 may acquire, from the storage unit 310, for example, a copy vector of a processing target sub-block included in information according to the coding mode. As described above, the motion vector in the prediction vector mode is determined to be the processing target sub-block before the coding mode change determination unit 418 executes the change of the coding mode of the sub-block in the coding block. A copy vector may be set.

  Further, the inter coding mode changing unit 415 selects a prediction vector based on selection information (for example, syntax mvp_l0_flag and mvp_l1_flag in HEVC) in order to encode a motion vector in the prediction vector mode. In one embodiment, the selection information may be predetermined information. For example, the selection information may be set to mvp_lX_flag = 0 that makes it easy to calculate a prediction vector candidate list (for example, mvpListLX (here, X = 0 or 1) in HEVC) generated for derivation of a prediction vector. . Alternatively, the inter coding mode changing unit 415 executes a process of creating a prediction vector candidate list (mvpListLX), selects mvp_lX_flag so that the cost of coding the difference vector between the prediction vector and the motion vector is reduced. Also good. For example, as described above, when the encoding mode is changed from the copy mode to the prediction vector mode, the motion vector and the prediction vector are set, so that data can be appropriately decoded in the transmission destination video decoding device. Is possible.

  The encoding mode change determination unit 418 determines whether to change the encoding mode of the sub-block to be processed. In the first embodiment, the encoding mode determination unit 412 has completed the determination as to whether to use an intra encoding mode other than IPCM or an inter encoding mode. Therefore, in the first embodiment, the encoding mode change determination unit 418 may determine whether or not to change the encoding mode to IPCM. As described above, for example, in HEVC, a limit value is provided for the code amount per CTU based on the code amount of IPCM. By normalizing the limit value per CTU in units of sub-blocks, It is possible to set a threshold value representing the code amount limitation in Then, the encoding mode change determination unit 418 determines whether or not the code amount of the processing target sub-block exceeds a set threshold value, and the code amount of the processing target sub-block exceeds the threshold value The coding mode of the sub-block to be processed may be changed to IPCM. Also, when the encoding mode change determination unit 418 changes the encoding mode of the sub-block to be processed to IPCM, the intra-block position storage unit 417 displays information indicating the position of the sub-block whose encoding mode has been changed to IPCM. Output to. Then, the encoding mode change determination unit 418 outputs the data of the sub-block that has been processed to the entropy encoding unit 404.

  The intra block position storage unit 417 stores information indicating the position of the sub block changed to the IPCM input from the encoding mode change determination unit 418. The intra block position storage unit 417 may store, for example, the position of the sub block changed to the IPCM as a bitmap using a flag.

  The entropy encoding unit 404 performs entropy encoding on the sub-block data input from the encoding mode change determination unit 418, and outputs a stream. Note that entropy coding may be a method (variable length coding) in which a variable length code is assigned according to the appearance frequency of a symbol.

  Subsequently, an operation flow of a moving image encoding process according to the first embodiment executed by the functional block configuration illustrated above will be exemplified. FIG. 5 is a diagram illustrating an operation flow of a moving image encoding process according to the first embodiment. The operation flow of the moving image encoding process in FIG. 5 may be started when, for example, information instructing encoding of a moving image is input to the control unit 300 in the moving image encoding device 30.

  In step 501 (hereinafter, “step” is described as “S”, for example, expressed as S501), the encoding mode determination unit 412 of the moving image encoding device 30 converts the picture in the moving image into a plurality of encoded blocks. To divide. For example, in HVEC, the control unit 300 may divide a picture into a plurality of CTUs. The subsequent processes from S502 to S519 are repetitive processes executed for each of the plurality of divided encoded blocks.

  In S503, the encoding mode determination unit 412 further divides the encoding block to be processed into a plurality of sub-blocks. For example, in HVEC, the encoding mode determination unit 412 may divide a CTU into a plurality of CUs. The subsequent processes from S504 to S506 are repetitive processes executed for each of the divided sub-blocks. In S505, the encoding mode determination unit 412 determines the encoding mode of the processing target sub-block.

  The determination of the encoding mode may be performed based on the encoding cost. For example, the determination may be performed by comparing the pixel difference absolute value sum: SAD = Σ | OrgPixel-PredPixel |. Here, SAD is an abbreviation for Sum of Absolute Difference. OrgPixel may be a pixel included in a block of an original image, and PredPixel may be a pixel included in a block of a predicted image determined by the HEVC standard, for example. However, the coding cost used in the embodiment is not limited to SAD, and may be, for example, an absolute value sum SATD after Hadamard transform of the difference pixel. SATD is an abbreviation for Sum of Absolute Hadamard Transformed Difference.

  As an example, in the case of the cost of inter coding, if the amount of information for coding the difference vector: MVD = (predicted vector-motion vector) is MVDCost in the prediction vector mode, the overall cost is Cost = SAD. + Can be expressed as λ * MVDCost. Here, λ is a scaler that adjusts the balance between SAD and MVDCost.

  Further, for example, in the case of the cost of intra coding, if the cost for coding in the intra prediction mode such as the cost for coding information on the intra prediction mode is PRDCost, the total cost is Cost = SAD + λ * It can be expressed as PRDCost. Here, λ is a scaler that adjusts the balance between SAD and PRDCost.

  Then, the encoding mode determination unit 412 compares the intra encoding cost of the processing target sub-block with the inter encoding cost, and determines the encoding mode by selecting the one with the lower cost. Good. Further, the motion vector used for calculating the cost of the inter coding mode may be determined by the motion vector calculation unit 411 using a block matching technique (motion vector search) as described above, for example. The motion vector may be determined for each PU in the CU. Note that MVDCost and PRDCost need not be used in determining the encoding mode of the sub-block to be processed. For example, the encoding mode may be determined by comparing SADs.

  Also, the coding mode determination unit 412 may determine the prediction mode of the processing target sub-block in S505. For example, the prediction mode may include a prediction vector mode and a copy mode if the prediction mode is an inter coding mode. As described above, the prediction vector mode may be a mode in which, for example, a motion vector of an encoded peripheral block is used as a prediction vector, and a difference vector from the motion vector obtained in the processing target sub-block is encoded. . The copy mode may be a mode for copying the motion vectors of the encoded peripheral sub-blocks. For example, the encoding mode determination unit 412 may obtain the SAD of the predicted image block predicted from the motion vector obtained by the motion vector search and the block to be processed as the cost of the predicted vector mode. Also, the encoding mode determination unit 412 generates a list of motion vector candidates (for example, a merge candidate list) derived from the motion vectors of the encoded peripheral sub-blocks. Then, the encoding mode determination unit 412 may obtain the SAD of the predicted image block predicted using the motion vector selected from the candidates and the processing target block as the cost of the copy mode. The encoding mode determination unit 412 may determine the prediction mode by comparing the cost of the prediction vector mode with the cost of the copy mode and selecting the one with the lower cost. Note that the encoding mode determination unit 412 may determine an index for designating a motion vector used in the copy mode from the list of motion vector candidates when calculating the cost of the copy mode. Further, the prediction mode may be determined for each PU in the CU.

  In S506, when all the subblocks obtained by dividing in S503 are not processed in S504 to S506, the flow returns to S503, and the control unit 300 repeats the process with the next subblock as a processing target. On the other hand, in S506, when the processes of S504 to S506 are executed for all the sub-blocks obtained by the division in S503, the flow proceeds to S507.

  The subsequent processing from S507 to S518 is, for example, repetitive processing executed for each of a plurality of sub-blocks (for example, CU) divided in S503. Further, the processing from S508 to S513 is an iterative process that is executed for each sub-block (for example, PU) of a unit that determines the prediction mode in the processing-target sub-block (for example, CU), for example. . In step S509, the copy mode determination unit 414 determines whether the processing target sub-block (for example, PU) is in the copy mode. If the sub-block to be processed is not in the copy mode (No in S509), the flow proceeds to S510. On the other hand, when the sub-block to be processed is in the copy mode (Yes in S509), the flow proceeds to S511.

  In S511, the adjacent sub-block determining unit 416 determines whether there is a sub-block (for example, CU) whose encoding mode is changed from the inter encoding mode to the IPCM in the sub-block adjacent to the processing target sub-block. . As will be described later in S517, when the coding mode of a certain subblock is changed from the inter coding mode to the IPCM, information about the changed subblock position is stored in the intra block position storage unit. 417 is stored. Therefore, the adjacent subblock determination unit 416 refers to the intra block position storage unit 417 and determines whether there is a subblock that has been changed from the inter coding mode to the IPCM in a subblock adjacent to the processing target subblock. You may judge. If there is no sub-block changed from the inter coding mode to the IPCM in the sub-block adjacent to the processing target sub-block (No in S511), the flow proceeds to S513. On the other hand, when there is a sub-block that has been changed from the inter coding mode to the IPCM in a sub-block adjacent to the processing target sub-block (Yes in S511), the flow proceeds to S512.

  In S512, the inter coding mode changing unit 415 changes the prediction mode of the processing target sub-block from the copy mode to the prediction vector mode. Note that the motion vector used in the prediction vector mode is a copy vector for the processing target sub-block determined before the coding mode of the sub-block in the coding block to be encoded is changed in S516 described later. It's okay. For example, this copy vector is processed until the sub-block data to be processed input to the prediction error signal generation unit 401 is input to the copy mode determination unit 414. 408 may be stored. Then, in S512, the inter coding mode changing unit 415 may acquire the copy vector in the copy mode determined for the processing target sub-block from the decoded image storage unit 408. Alternatively, in another embodiment, in S505, the encoding mode determination unit 412 determines the encoding mode of the processing target sub-block and information according to the encoding mode, for example, the storage unit 310 may be stored. In step S512, the inter coding mode changing unit 415 may acquire, from the storage unit 310, for example, a copy vector of the processing target sub-block included in the information corresponding to the coding mode.

  Subsequently, in S510, the inter coding mode changing unit 415 calculates a prediction vector. For example, the inter coding mode changing unit 415 may select a prediction vector based on selection information (for example, syntax mvp_l0_flag and mvp_l1_flag in HEVC) in order to encode a motion vector. In one embodiment, the selection information may be predetermined information that is determined in advance. For example, the selection information may be set to mvp_lX_flag = 0 that makes it easy to calculate a prediction vector candidate list (for example, mvpListLX (here, X = 0 or 1) in HEVC) generated for derivation of a prediction vector. . Alternatively, in another embodiment, mvp_lX_flag may be selected so that the process of creating a prediction vector candidate list (mvpListLX) is executed and the cost of encoding the difference vector between the prediction vector and the motion vector is reduced. . For example, even when the encoding mode is changed from the copy mode to the prediction vector mode by performing the encoding using the motion vector and the selection information for selecting the prediction vector as described above, the transmission destination Data can be appropriately decoded in the moving picture decoding apparatus. In S510, for example, if the sub-block to be processed is an intra coding mode or IPCM, the prediction vector need not be calculated.

  In S513, is the control unit 300 performing the repetition processing from S508 to S513 on each of the sub-blocks (for example, PUs) for determining the prediction mode in the processing target sub-block (for example, CU)? Determine whether or not. In S513, when the iterative process from S508 to S513 is not executed for each of the sub-blocks (for example, PUs) for determining the prediction mode, the flow returns to S508, and the control unit 300 Repeat the process for the sub-block. On the other hand, in S513, when the iterative process from S508 to S513 is executed for each of the sub-blocks (for example, PUs) for determining the prediction mode, the flow proceeds to S514.

  In S514, the coding mode change determination unit 418 determines whether or not to change the coding mode of the processing target sub-block (for example, CU) to the intra coding mode. In the first embodiment, whether the encoding mode determination unit 412 uses an intra encoding mode other than the IPCM or an inter encoding mode for encoding the sub-block to be processed in S505. Judgment is being executed. Therefore, the encoding mode change determination unit 418 may determine whether or not to change the encoding mode to IPCM in S514.

  Further, this determination may be performed as follows. As described above, for example, in HEVC, a limit value is provided for the code amount per CTU based on the code amount of IPCM. Then, by normalizing the limit value per CTU in units of sub-blocks (for example, CU), it is possible to set a threshold value indicating the code amount limit in units of sub-blocks. Therefore, the encoding mode change determination unit 418 may determine whether or not the code amount of the sub-block to be processed exceeds a threshold representing the code amount limit in S514. For example, if the code amount of the sub-block to be processed does not exceed the threshold, the encoding mode change determination unit 418 determines No in S514, and the flow proceeds to S515. On the other hand, if the code amount of the sub-block to be processed exceeds the threshold, the encoding mode change determination unit 418 determines Yes in S514, and the flow proceeds to S516.

  In S516, the encoding mode change determination unit 418 changes the encoding mode of the processing target sub-block to IPCM. In S517, the coding mode change determination unit 418 stores information indicating the position of the sub-block whose coding mode is changed to the IPCM in the intra block position storage unit 417. For example, the intra block position storage unit 417 may store the position of the sub block changed to the IPCM as a bitmap using a flag. Subsequently, in S515, the entropy encoding unit 404 may perform entropy encoding on the processing target sub-block data input from the encoding mode change determination unit 418, and output a bitstream. Note that entropy coding may be a method (variable length coding) in which a variable length code is assigned according to the appearance frequency of a symbol.

  In S518, the control unit 300 determines whether or not iterative processing from S507 to S518 is being performed on each of a plurality of sub-blocks (for example, CUs) obtained by dividing in S503. In S518, when iterative processing from S507 to S518 is not executed for each of the plurality of subblocks obtained by dividing in S503, the flow returns to S507, and the control unit 300 determines that the next subblock The process is repeated for. On the other hand, in S518, when the iterative processing from S507 to S518 is executed for each of the plurality of sub-blocks obtained by dividing in S503, the flow proceeds to S519.

  In S519, the control unit 300 determines whether or not the iterative processing from S502 to S519 has been executed for all the encoded blocks obtained by the division in S501. If there is an unprocessed encoded block, the flow returns to S502. On the other hand, when the iterative process from S502 to S519 is executed for all the encoded blocks obtained by the division in S501, the operation flow ends.

  In the operation flow of FIG. 5 described above, for example, the processing of S504 to S506 may be executed by the determination unit 301 included in the control unit 300. The processing of S509 to S512 may be executed by the changing unit 302 included in the control unit 300.

  FIG. 6 is a diagram exemplifying the change to the IPCM of the coding mode of the sub-block to be processed and the change to the prediction vector mode of the surrounding sub-block accompanying the above-described change. 6A, the processing target sub-block (for example, CU) in FIG. 2A is changed to IPCM. In this case, in the adjacent subblock (for example, PU) in the merge mode indicated by the oblique line 601 in FIG. 6A, the subblock changed to the IPCM may be referred to as a motion vector copy source in the merge mode. There is. In this case, according to the above operation flow, as shown in FIG. 6B, adjacent sub-blocks are changed from the merge mode to the AMVP mode. Further, a copy vector that is scheduled to be used in the merge mode is explicitly encoded as a motion vector in the AMVP mode. Therefore, according to the first embodiment, when the coding mode of a certain subblock in the coding block is changed from the inter coding mode to the IPCM, the vector copy relationship that occurs in the subblocks in the adjacent copy mode Inconsistency can be solved. In addition, since inconsistencies in vector copy relations are resolved in adjacent copy mode sub-blocks, errors may be chained to adjacent copy mode sub-blocks around adjacent copy mode sub-blocks. Can be avoided.

  Further, in the above-described embodiment, when the prediction mode is changed to the prediction vector mode, the motion vector is determined before the sub-block in the coding block is changed from the inter coding mode to the intra coding mode. A copy vector is used. Therefore, the predicted image to be encoded does not change. Further, since the copy vector that has already been determined is used as the motion vector, for example, it is not necessary to perform motion search again, and the amount of calculation can be suppressed. Furthermore, in the above embodiment, when the sub-block prediction mode is changed from the copy mode to the prediction vector mode, a prediction vector is calculated in S510. When the prediction mode is changed from the copy mode to the prediction vector mode, the motion vector used in the prediction vector mode is specified and the prediction vector is calculated. Decoding can be performed appropriately in the decoding device.

  Moreover, in the calculation of a prediction vector, a prediction vector candidate is derived and a prediction vector is selected from the candidates. Here, for example, by using predetermined information as selection information for selecting a prediction vector, it is not necessary to derive all prediction vector candidates, and it is possible to reduce the amount of calculation for determining a prediction vector. .

  Alternatively, in another embodiment, a prediction vector candidate may be derived in S510, and the prediction vector may be selected so that the cost for encoding the difference vector between the prediction vector and the motion vector is reduced from among the candidates. . By selecting the prediction vector in this way, the code amount in the prediction vector mode can be reduced. Note that the calculation for deriving the prediction vector in the prediction vector mode has a smaller amount of calculation than the calculation for deriving the copy vector in the copy mode. Therefore, for example, when the coding mode of a certain sub-block in the coding block is changed from the inter coding mode to the IPCM, compared with the case where the copy vector is recalculated for the sub-blocks in the surrounding copy mode. Thus, the calculation amount can be suppressed.

  In the above-described embodiment, the example in which it is determined in S511 whether or not there is a sub-block that has been changed from the inter coding mode to the IPCM in a sub-block adjacent to the processing target sub-block. However, the embodiment is not limited to this. For example, in the HEVC, sub-blocks that are referred to in order to generate a motion vector candidate list for a processing target sub-block (for example, PU) are defined in the copy mode. FIG. 7 is a diagram illustrating sub-blocks that are referred to in HEVC for generating a motion vector candidate list in the copy mode. In FIG. 7, A0, A1, B0, B1, and B2 are sub-blocks that are referred to in order to generate a motion vector candidate list. Therefore, in S511, the adjacent subblock determination unit 416 determines whether there is a subblock that has been changed to the IPCM in a block that is a reference target in the copy mode among subblocks adjacent to the processing target subblock. It may be determined. Thereby, for example, in the process of S512, it is possible to reduce the number of sub-blocks for which the prediction mode is changed from the copy mode to the prediction vector mode, and thus it is possible to further reduce the processing load according to the embodiment.

<Second Embodiment>
In the first embodiment, the encoding mode of each of the plurality of sub-blocks in the encoding block is determined to be the inter encoding mode or the intra encoding mode, and then the encoding mode is changed to IPCM. An example of. However, the embodiment is not limited to this. For example, the moving image encoding apparatus 30 first determines the prediction mode and the motion vector in the inter encoding mode by setting the encoding mode of all the sub-blocks in the encoding block to be processed as the inter encoding mode. Thereafter, the moving image encoding device 30 may, for example, change from the inter encoding mode to the intra encoding mode based on a comparison between the pixel difference absolute value sum in the inter encoding mode and the pixel difference absolute value sum in the intra encoding mode. It may be determined whether or not to change. Even in such a case, if there is a copy mode sub-block around the sub-block changed from the inter coding mode to the intra coding mode, copying is performed in the copy mode due to the change of the coding mode. The motion vector to be changed may change.

  Therefore, in the second embodiment, when there is a copy mode sub-block around the sub-block changed to the intra coding mode, the video encoding device 30 sets the copy mode sub-block to the prediction vector mode. change. Hereinafter, the second embodiment will be described.

  Note that, also in the second embodiment, the video encoding device 30 may have the functional block configuration shown in FIGS. 3 and 4, for example. However, in the second embodiment, for example, the functional block configuration may perform an operation different from that in the first embodiment.

  For example, in the second embodiment, the encoding mode determination unit 412 determines all the subblocks in the processing target encoding block as the inter encoding mode, and determines the motion vector and the prediction mode of each subblock. . As described above, when all the sub-blocks in the coding block to be coded are set to the inter coding mode and the motion vector and the prediction mode of each sub-block are determined, these processes can be performed in parallel. Is possible. For example, executing the determination of the motion vector and the prediction mode by parallel processing is useful for securing real-time properties.

  Also, the coding mode change determination unit 418 determines whether or not to change the processing target sub-block from the inter coding mode to the intra coding mode. This determination may be performed based on the coding cost. For example, the determination may be performed by comparing the pixel difference absolute value sum: SAD = Σ | OrgPixel−PredPixel |. However, the coding cost used in the embodiment is not limited to SAD, and may be, for example, an absolute value sum SATD after Hadamard transform of the difference pixel. For example, the coding mode change determination unit 418 compares the intra coding cost of the sub-block to be processed with the inter coding cost, and performs coding when the intra coding cost is lower. The mode may be changed to the intra coding mode.

  Furthermore, the encoding mode change determination unit 418 may determine whether or not to change the encoding mode to IPCM, which is one of the intra encoding modes, as in the first embodiment. This determination may be performed using a threshold normalized on a sub-block (eg, CU) basis, as described above.

  The intra block position storage unit 417 stores information indicating the position of the sub block when the encoding mode change determination unit 418 changes the processing target sub block to the intra coding mode (including IPCM). Good.

  The copy mode determination unit 414 determines whether the encoding mode of the processing target sub-block is the copy mode. When the encoding mode of the subblock to be processed is the copy mode, the copy mode determination unit 414 inputs information indicating the position of the subblock determined to be the copy mode to the adjacent subblock determination unit 416. The adjacent sub-block determining unit 416 determines whether the sub-block changed to the intra coding mode (including IPCM) is included in the sub-block adjacent to the copy mode sub-block to which the position information is input. To do. This determination may be performed based on, for example, information indicating the position of the sub block changed to the intra coding mode (including IPCM) stored in the intra block position storage unit 417.

  The inter coding mode changing unit 415 handles the processing when the sub block to be processed is in the copy mode and the adjacent sub block includes a sub block changed to the intra coding mode (including IPCM). Are changed to the prediction vector mode. Further, the inter coding mode changing unit 415, for example, uses the copy vector determined in the copy mode for the processing target sub-block before the adjacent sub-block is changed to the intra coding mode, as a prediction vector. Used as a motion vector in mode. For example, for the motion vector used in the prediction vector mode, when the encoding mode determination unit 412 determines the encoding mode with all the sub blocks as the inter encoding mode, the copy determined for the processing target sub block is used. Vectors may be used.

  Subsequently, an operation flow of a moving image encoding process according to the second embodiment executed by the functional block configuration exemplified above will be exemplified. FIG. 8 is a diagram illustrating an operation flow of a moving image encoding process according to the second embodiment. The operation flow of the moving image encoding process in FIG. 8 may be started when, for example, information instructing encoding of a moving image is input to the control unit 300 in the moving image encoding device 30.

  The operation flow in FIG. 8 is an operation flow corresponding to the operation flow in FIG. 5 in the first embodiment. For example, in the processes of S801 to S804, S806 to S810, S813, S815, S818, and S819, the same processes as the processes of S501 to S504, S506 to S510, S513, S515, S518, and S519 in FIG. Good. However, for example, in the following processing, processing different from that in FIG. 5 may be executed.

  In S805, the coding mode determination unit 412 determines the motion vector and the prediction mode for the processing target sub-block, with the coding mode of all the sub-blocks included in the coding block as the inter-coding mode.

  In S811, the adjacent subblock determination unit 416 is changed from the inter coding mode to the intra coding mode (including IPCM) for the subblock (for example, CU) adjacent to the processing target subblock (for example, PU). It may be determined whether there is another sub-block. As will be described later in S817, when the coding mode of a certain subblock is changed from the inter coding mode to the intra coding mode (including IPCM), the position of the changed subblock is changed. Information is stored in the intra block position storage unit 417. Therefore, the adjacent subblock determination unit 416 refers to the intra block position storage unit 417 and determines whether there is a subblock that has been changed from the inter coding mode to the intra coding mode (including IPCM) in the adjacent subblock. It may be determined whether or not. If there is no subblock changed from the inter coding mode to the intra coding mode (including IPCM) in the subblock adjacent to the processing target subblock (S811 is No), the flow proceeds to S813. On the other hand, when there is a subblock that has been changed from the inter coding mode to the intra coding mode (including IPCM) in the subblock adjacent to the processing target subblock (Yes in S811), the flow proceeds to S812. In step S812, the inter coding mode changing unit 415 changes the prediction mode of the processing target sub-block (for example, PU) from the copy mode to the prediction vector mode.

  In S814, the coding mode change determination unit 418 determines whether or not to change the coding mode of the processing target sub-block (for example, CU) to the intra coding mode. In the second embodiment, as described above, in S805, all the sub-blocks in the coding block to be coded are processed in the inter coding mode. Therefore, the coding mode change determination unit 418 may determine whether or not to change the coding mode to the intra coding mode in S814.

  This determination may be performed based on the coding cost. For example, the determination may be performed by comparing the pixel difference absolute value sum: SAD = Σ | OrgPixel−PredPixel |. However, the coding cost used in the embodiment is not limited to SAD, and may be, for example, an absolute value sum SATD after Hadamard transform of the difference pixel. For example, the coding mode change determination unit 418 compares the intra coding cost of the sub-block to be processed with the inter coding cost, and performs coding when the intra coding cost is lower. The mode may be changed to the intra coding mode.

  Furthermore, in S814, the coding mode change determination unit 418 may determine whether or not to change the coding mode to an IPCM that is one of the intra coding modes, as in the first embodiment. This determination may be performed using a threshold normalized on a sub-block (eg, CU) basis, as described above.

  If the inter coding mode is not changed to the intra coding mode (including IPCM) in S814 (No in S814), the flow proceeds to S815. On the other hand, when the inter coding mode is changed to the intra coding mode (including IPCM) (Yes in S814), the flow proceeds to S816. In S816, the coding mode change determination unit 418 changes the coding mode to the intra coding mode (including IPCM). In S817, the coding mode change determination unit 418 stores information indicating the position of the sub-block whose coding mode is changed to the intra coding mode (including IPCM) in the intra block position storage unit 417.

  In the operation flow of FIG. 8, for example, the processing of S804 to S806 may be executed by the determination unit 301 included in the control unit 300. Further, the processing of S809 to S812 may be executed by the changing unit 302 included in the control unit 300.

  As described above, in the second embodiment, the prediction mode and the motion vector in the inter coding mode are determined with the coding mode of all the sub-blocks in the coding block to be processed as the inter coding mode. doing. After that, when the coding mode of a certain subblock (for example, CU) is changed to the intra coding mode (including IPCM), the subblock (for example, PU) in the copy mode adjacent to the subblock is changed. The prediction vector mode has been changed. Therefore, the second embodiment has the effect of the first embodiment. Further, according to the second embodiment, even when a certain sub-block is changed from the inter-coding mode to the intra-coding mode after the coding mode is determined, the sub-blocks in the copy mode around the sub-block Inconsistency in the vector copy relationship that occurs in

  Although several embodiments have been exemplified above, the embodiments are not limited to the above-described embodiments. For example, in the above example, the moving image coding by HEVC has been described as an example, but the embodiment is not limited to this. For example, H.M. In other moving image coding systems such as H.264 and MPEG-2, a coding mode of a plurality of sub-blocks (for example, a macro block) in a certain block is determined in a lump, and then the coding mode is changed. If so, embodiments can be applied as well.

  Further, for example, the operation flows of FIGS. 5 and 8 are examples, and the embodiment is not limited to this. For example, in the operation flows of FIGS. 5 and 8, the order of processing may be changed, if possible, may include additional processing, or some processing may be omitted. Good.

  Further, the moving image encoding apparatus 30 according to an embodiment can be implemented as a hardware circuit, and can be realized by using, for example, an information processing apparatus (computer) 900 as shown in FIG.

  9 includes a processor 901, a memory 902, an input device 903, an output device 904, a storage device 905, a medium drive device 906, and a network connection device 907, and these components are connected to each other by a bus 909. Has been.

  The memory 902 is, for example, a semiconductor memory such as a read only memory (ROM), a random access memory (RAM), or a flash memory. The memory 902 stores a moving image encoding program used for moving image encoding processing and data such as moving image data. The storage unit 310 in FIG. 3, the decoded image storage unit 408, and the intra block position storage unit 417 in FIG. 4 may be, for example, the memory 902.

  The processor 901 may be, for example, a Central Processing Unit (CPU). 3 may be the processor 901, for example. The processor 901 may operate as the determination unit 301 and the change unit 302 in FIG. 3, for example, by executing a moving image encoding program using the memory 902. Further, the processor 901, for example, the prediction error signal generation unit 401, the orthogonal transformation unit 402, the quantization unit 403, the entropy coding unit 404, the inverse quantization unit 405, the inverse orthogonal transformation unit 406, and the decoded image generation illustrated in FIG. The unit 407 may operate. The processor 901 may operate as, for example, the intra prediction image generation unit 409, the inter prediction image generation unit 410, the motion vector calculation unit 411, the encoding mode determination unit 412, and the prediction image selection unit 413. The processor 901 may operate as the copy mode determination unit 414, the inter coding mode change unit 415, the adjacent subblock determination unit 416, and the coding mode change determination unit 418.

  The input device 903 is, for example, a keyboard, a pointing device, or the like, and is used for inputting an instruction or information from a user or an operator. The output device 904 is, for example, a display device, a printer, a speaker, or the like, and is used to output an inquiry to a user or an operator or a processing result.

  The storage device 905 is, for example, a magnetic disk device, an optical disk device, a magneto-optical disk device, a tape device, or the like. The storage device 905 may be a hard disk drive. The information processing apparatus 900 can store a moving image encoding program and data such as moving image data in the storage device 905 and load them into the memory 902 for use.

  The medium driving device 906 drives the portable recording medium 910 and accesses the recorded contents. The portable recording medium 910 is a memory device, a flexible disk, an optical disk, a magneto-optical disk, or the like. The portable recording medium 910 may be a Compact Disk Read Only Memory (CD-ROM), a Digital Versatile Disk (DVD), or a Universal Serial Bus (USB) memory. A user or an operator can store a moving image encoding program and data such as moving image data in the portable recording medium 910 and load them into the memory 902 for use.

  As described above, a computer-readable recording medium that stores a moving image encoding program and data such as moving image data is a physical (non-storage medium) such as the memory 902, the storage device 905, or the portable recording medium 910. A temporary recording medium.

  The network connection device 907 is a communication interface that is connected to a communication network such as a local area network (LAN) or the Internet and performs data conversion accompanying communication. The network connection device 907 can also be used as an output interface that outputs an encoded stream to the video decoding device. The information processing apparatus 900 can also receive a moving image encoding program and data such as moving image data from an external device via the network connection device 907 and load them into the memory 902 for use.

  The hardware configuration of the information processing apparatus 900 illustrated in FIG. 9 is an exemplification, and the embodiment is not limited to this. For example, in the information processing apparatus 900, some components may be omitted depending on the application and conditions. For example, when the interface with the user or the operator is not used, the input device 903 and the output device 904 may be omitted. Further, when the information processing apparatus 900 does not access the portable recording medium 910, the medium driving apparatus 906 may be omitted.

  Further, for example, in another embodiment, a part or all of the functions of the control unit 300 of the above-described video encoding device 30 may be implemented as hardware such as FPGA and SoC. Note that FPGA is an abbreviation for Field Programmable Gate Array. SoC is an abbreviation for System-on-a-chip. Furthermore, when the processes included in the above-described operation flow are possible, a part of the processes may be executed in parallel processing. For example, a part of the processes may be executed by another processor. By executing in parallel processing, the real-time property of the moving image encoding processing can be ensured.

  Several embodiments, including those described above, will be understood by those skilled in the art as encompassing various variations and alternatives of the embodiments described above. For example, various embodiments may be embodied by modifying components. Various embodiments may be implemented by appropriately combining a plurality of components disclosed in the above-described embodiments. Further, various embodiments may be implemented by deleting or replacing some components from all the components shown in the embodiments, or adding some components to the components shown in the embodiments. May be.

30 moving image coding apparatus 300 control unit 301 determination unit 302 change unit 310 storage unit 401 prediction error signal generation unit 402 orthogonal transform unit 403 quantization unit 404 entropy coding unit 405 inverse quantization unit 406 inverse orthogonal transform unit 407 decoded image Generation unit 408 Decoded image storage unit 409 Intra prediction image generation unit 410 Inter prediction image generation unit 411 Motion vector calculation unit 412 Coding mode determination unit 413 Prediction image selection unit 414 Copy mode determination unit 415 Inter coding mode change unit 416 Adjacent sub Block determination unit 417 Intra block position storage unit 418 Coding mode change determination unit 900 Information processing device 901 Processor 902 Memory 903 Input device 904 Output device 905 Storage device 906 Medium drive device 907 Network connection device 909 Bus 910 Portable Recording media

Claims (7)

Determining a coding mode to be applied to each of a plurality of blocks in the image, and determining a prediction mode to be applied to each of the plurality of blocks;
After the determining unit determines that the encoding mode of the first block among the plurality of blocks is the inter encoding mode, the encoding mode of the first block is changed to the intra encoding mode, and the plurality of blocks When the prediction mode of the second block adjacent to the first block among the blocks is determined to be the copy mode for copying the motion vector of the adjacent block, the prediction mode of the second block is changed from the copy mode. A changing unit for changing to a prediction vector mode for encoding a difference vector between a motion vector and a prediction vector;
A moving picture encoding apparatus.   When the encoding unit of the second block is changed from the copy mode to the prediction vector mode, the changing unit determines whether the determining unit applies the motion vector used in the prediction vector mode to the second block. The moving image encoding apparatus according to claim 1, wherein a motion vector to be copied in the copy mode determined in the above is used.   The moving image encoding apparatus according to claim 1, wherein the changing unit sets predetermined information as selection information for selecting the prediction vector used in the prediction vector mode.   The changing unit derives the prediction vector candidate used in the prediction vector mode for the second block, and the cost for encoding the difference vector for the second block out of the candidates is small. The moving picture coding apparatus according to claim 1, wherein the prediction vector is selected so that   The changing unit changes the prediction mode of the second block from the copy mode to the prediction vector mode, and the position of the first block relative to the second block moves in the copy mode. 5. The moving picture encoding apparatus according to claim 1, wherein the moving picture encoding apparatus is executed when the position is defined as a vector reference destination. Determining a coding mode applied to each of a plurality of blocks in the image, and determining a prediction mode applied to each of the plurality of blocks;
After the coding mode of the first block among the plurality of blocks is determined to be the inter coding mode, the coding mode of the first block is changed to the intra coding mode, When the prediction mode of the second block adjacent to the first block is determined as the copy mode for copying the motion vector of the adjacent block, the prediction mode of the second block is changed from the copy mode to the motion vector. Changing to a prediction vector mode for encoding a difference vector with a prediction vector;
A moving picture encoding method executed by a computer. Determining a coding mode applied to each of the plurality of blocks in the image, and determining a prediction mode applied to each of the plurality of blocks;
After the coding mode of the first block among the plurality of blocks is determined to be the inter coding mode, the coding mode of the first block is changed to the intra coding mode, When the prediction mode of the second block adjacent to the first block is determined as the copy mode for copying the motion vector of the adjacent block, the prediction mode of the second block is changed from the copy mode to the motion vector. Change to the prediction vector mode that encodes the difference vector with the prediction vector,
A moving image encoding program for causing a computer to execute processing.

Images