WO2018173895A1

WO2018173895A1 - Predicted-image generation device, video decoding device, and video coding device

Info

Publication number: WO2018173895A1
Application number: PCT/JP2018/010058
Authority: WO
Inventors: 知宏猪飼; 友子青野
Original assignee: シャープ株式会社
Priority date: 2017-03-21
Filing date: 2018-03-14
Publication date: 2018-09-27

Abstract

The purpose of the present invention is to effectively utilize the matching cost of a motion-compensated image. An inter prediction parameter decoding unit (303) is provided with: a BTM processing unit (3038) which performs matching using BTM or TM and derives a matching cost of a motion-compensated image; and a weight prediction unit (3094) which sets a weighting coefficient for each of motion-compensated images in accordance with the matching cost, and derives the predicted image using the weighted motion-compensated images.

Description

Predicted image generation device, moving image decoding device, and moving image encoding device

Embodiments described herein relate generally to a predicted image generation device, a moving image decoding device, and a moving image encoding device.

In order to efficiently transmit or record a moving image, a moving image encoding device that generates encoded data by encoding the moving image, and a moving image that generates a decoded image by decoding the encoded data An image decoding device is used.

Specific examples of the moving image encoding method include a method proposed in H.264 / AVC and HEVC (High-Efficiency Video Coding).

In such a moving image coding system, an image (picture) constituting a moving image is a slice obtained by dividing the image, a coding unit obtained by dividing the slice (coding unit (Coding Unit : CU)), and a hierarchical structure consisting of a prediction unit (PU) and a transform unit (TU) that are obtained by dividing a coding unit. Decrypted.

In such a moving image coding method, a predicted image is usually generated based on a local decoded image obtained by encoding / decoding an input image, and the predicted image is generated from the input image (original image). A prediction residual obtained by subtraction (sometimes referred to as “difference image” or “residual image”) is encoded. Examples of the method for generating a predicted image include inter-screen prediction (inter prediction) and intra-screen prediction (intra prediction).

Further, Non-Patent Document 1 using matching motion deriving technology (template matching (TM) and bilateral matching (BM)) for deriving a motion vector on the decoder side can be cited as a moving image encoding and decoding technology in recent years.

In Non-Patent Document 2, bilateral template matching (BTM) that corrects a motion vector by matching is described in the name of Decoder-side motion vector refinement (DMVR) as a similar technique to the matching motion derivation technology. .

The configurations disclosed in

Non-Patent Documents

1 and 2 above show a configuration that effectively uses the matching cost of prediction image candidates (motion compensated images and interpolated images) derived in bilateral template matching or template matching. Absent.

The present invention is to provide a predicted image generation device, a moving image decoding device, and a moving image encoding device that can effectively use the matching cost of the predicted image candidate (motion compensation image).

In order to solve the above problem, a predicted image generation apparatus according to an aspect of the present invention is a predicted image generation apparatus that generates a predicted image used for encoding or decoding of a moving image, and includes a bilateral template matching (BTM ) Or template matching (TM) is used to perform matching processing, a matching processing unit for deriving the cost of the predicted image candidate derived from the template in a plurality of reference pictures and the template, and prediction for each reference picture A prediction unit for deriving a prediction image from a plurality of prediction image candidates. The prediction unit sets a weighting coefficient for each prediction image candidate according to the cost of the template and the prediction image candidate. Then, the predicted image is derived using the weighted predicted image candidates.

In order to solve the above problem, a predicted image generation apparatus according to an aspect of the present invention is a predicted image generation apparatus that generates a predicted image used for encoding or decoding of a moving image, and includes a bilateral template matching (BTM ) And a matching processing unit for deriving the cost of predicted image candidates predicted from templates in a plurality of reference pictures and the template, and a gradient correction term for bilateral template matching (BTM) A motion compensation unit that performs motion compensation by bi-predictive gradient change (BIO) prediction that performs motion compensation by applying the motion compensation unit, and the motion compensation unit predicts the predicted image candidate predicted from a template in a plurality of reference pictures. ON / OFF of BIO prediction is controlled according to the ratio of the cost to the template.

According to the above configuration, the bilateral template matching or the matching cost of the predicted image candidate derived in the template matching can be effectively used for deriving the predicted image.

It is a figure which shows the hierarchical structure of the data of the encoding stream which concerns on this embodiment. It is a figure which shows the pattern of PU division | segmentation mode. (A) to (h) respectively show the partition shapes when the PU partitioning modes are 2Nx2N, 2NxN, 2NxnU, 2NxnD, Nx2N, nLx2N, nRx2N, and NxN. (A) And (b) is a conceptual diagram which shows an example of a reference picture and a reference picture list. It is a block diagram which shows the structure of the image coding apparatus which concerns on this embodiment. It is the schematic which shows the structure of the image decoding apparatus which concerns on this embodiment. It is the schematic which shows the structure of the inter estimated image generation part of the image coding apparatus which concerns on this embodiment. It is the schematic which shows the structure of the merge prediction parameter derivation | leading-out part which concerns on this embodiment. It is the schematic which shows the structure of the AMVP prediction parameter derivation | leading-out part which concerns on this embodiment. It is a flowchart which shows the operation | movement of the motion vector decoding process of the image decoding apparatus which concerns on this embodiment. It is the schematic which shows the structure of the inter prediction parameter encoding part of the image coding apparatus which concerns on this embodiment. It is the schematic which shows the structure of the inter estimated image generation part which concerns on this embodiment. It is the schematic which shows the structure of the inter prediction parameter decoding part which concerns on this embodiment. It is a figure which shows the example which derives | leads-out the motion vector spMvLX [xi] [yi] of each subblock which comprises PU (horizontal width nPbW) which is the object which estimates a motion vector. (A) is a figure for demonstrating bilateral matching (Bilateral (matching) matching). (B) is a figure for demonstrating template matching (Template | matching). It is a flowchart figure which shows the outline | summary of a motion prediction mode determination flow. (A) is a figure which shows the relationship between the reference image in a BTM process, and a template, (b) is a figure which shows the flow of a process. It is a figure for demonstrating the detail of the template in a BTM process. (A), (b) is a figure which shows the outline | summary of template matching. (A), (b) is a figure which shows the outline | summary of bilateral matching. It is a flowchart figure which shows the flow of the motion vector derivation | leading-out process in matching motion derivation mode (template matching, bilateral matching). (A) to (b) is a diagram for explaining a motion vector deriving process in the matching motion deriving mode. It is a block diagram which shows an example of the principal part structure of the inter estimated image generation part in other embodiment of this invention. It is a figure which shows the example of the weight which the weight estimation part in other embodiment of this invention derives. It is a block diagram which shows another example of the principal part structure of the inter estimated image generation part in other embodiment of this invention. It is a figure which shows an example of the pseudo code which shows the process of the motion compensation part in other embodiment of this invention. It is the figure shown about the structure of the transmitter which mounts the image coding apparatus which concerns on this embodiment, and the receiver which mounts an image decoding apparatus. (A) shows a transmission device equipped with an image encoding device, and (b) shows a reception device equipped with an image decoding device. It is the figure shown about the structure of the recording device carrying the image coding apparatus which concerns on this embodiment, and the reproducing | regenerating apparatus carrying an image decoding apparatus. (A) shows a recording device equipped with an image encoding device, and (b) shows a playback device equipped with an image decoding device. It is the schematic which shows the structure of the image transmission system which concerns on this embodiment.

Embodiment 1
Hereinafter, embodiments of the present invention will be described with reference to the drawings.

FIG. 28 is a schematic diagram showing the configuration of the image transmission system 1 according to the present embodiment.

The image transmission system 1 is a system that transmits a code obtained by encoding an encoding target image, decodes the transmitted code, and displays an image. The image transmission system 1 includes an image encoding device (moving image encoding device) 11, a network 21, an image decoding device (moving image decoding device) 31, and an image display device 41.

The image encoding device 11 receives an image T indicating a single layer image or a plurality of layers. A layer is a concept used to distinguish a plurality of pictures when there are one or more pictures constituting a certain time. For example, when the same picture is encoded with a plurality of layers having different image quality and resolution, scalable encoding is performed, and when a picture of a different viewpoint is encoded with a plurality of layers, view scalable encoding is performed. When prediction is performed between pictures of a plurality of layers (inter-layer prediction, inter-view prediction), encoding efficiency is greatly improved. Further, even when prediction is not performed (simultaneous casting), encoded data can be collected.

The network 21 transmits the encoded stream Te generated by the image encoding device 11 to the image decoding device 31. The network 21 is the Internet, a wide area network (WAN: Wide Area Network), a small network (LAN: Local Area Network), or a combination thereof. The network 21 is not necessarily limited to a bidirectional communication network, and may be a unidirectional communication network that transmits broadcast waves such as terrestrial digital broadcasting and satellite broadcasting. The network 21 may be replaced with a storage medium that records an encoded stream Te such as a DVD (Digital Versatile Disc) or a BD (Blue-ray Disc).

The image decoding device 31 decodes each of the encoded streams Te transmitted by the network 21, and generates one or a plurality of decoded images Td decoded.

The image display device 41 displays all or part of one or more decoded images Td generated by the image decoding device 31. The image display device 41 includes, for example, a display device such as a liquid crystal display or an organic EL (Electro-luminescence) display. In addition, in the spatial scalable coding and SNR scalable coding, when the image decoding device 31 and the image display device 41 have a high processing capability, a high-quality enhancement layer image is displayed and only a lower processing capability is provided. Displays a base layer image that does not require higher processing capability and display capability as an extension layer.

<Operator>
The operators used in this specification are described below.

>> is right bit shift, << is left bit shift, & is bitwise AND, | is bitwise OR, | = is sum operation (OR) with another condition.

X? Y: z is a ternary operator that takes y when x is true (non-zero) and takes z when x is false (0).

Clip3 (a, b, c) is a function that clips c to a value between a and b, but returns a if c <a, returns b if c> b, otherwise Is a function that returns c (where a <= b).

<Structure of encoded stream Te>
Prior to detailed description of the image encoding device 11 and the image decoding device 31 according to the present embodiment, a data structure of an encoded stream Te generated by the image encoding device 11 and decoded by the image decoding device 31 will be described. .

FIG. 1 is a diagram showing a hierarchical structure of data in the encoded stream Te. The encoded stream Te illustratively includes a sequence and a plurality of pictures constituting the sequence. (A) to (f) of FIG. 1 respectively show an encoded video sequence defining a sequence SEQ, an encoded picture defining a picture PICT, an encoded slice defining a slice S, and an encoded slice defining a slice data It is a figure which shows the coding unit (Coding | unit: CU) contained in the coding tree unit contained in data and coding slice data, and a coding tree unit.

(Encoded video sequence)
In the encoded video sequence, a set of data referred to by the image decoding device 31 for decoding the sequence SEQ to be processed is defined. As shown in FIG. 1A, the sequence SEQ includes a video parameter set (Video Parameter Set), a sequence parameter set SPS (Sequence Parameter Set), a picture parameter set PPS (Picture Parameter Set), a picture PICT, and an addition. Includes SEI (Supplemental Enhancement Information). Here, the value indicated after # indicates the layer ID. Although FIG. 1 shows an example in which encoded data of # 0 and # 1, that is, layer 0 and layer 1, exists, the type of layer and the number of layers are not dependent on this.

The video parameter set VPS is a set of encoding parameters common to a plurality of moving images, a plurality of layers included in the moving image, and encoding parameters related to individual layers in a moving image composed of a plurality of layers. A set is defined.

The sequence parameter set SPS defines a set of encoding parameters that the image decoding device 31 refers to in order to decode the target sequence. For example, the width and height of the picture are defined. A plurality of SPSs may exist. In that case, one of a plurality of SPSs is selected from the PPS.

In the picture parameter set PPS, a set of encoding parameters referred to by the image decoding device 31 in order to decode each picture in the target sequence is defined. For example, a quantization width reference value (pic_init_qp_minus26) used for picture decoding and a flag (weighted_pred_flag) indicating application of weighted prediction are included. There may be a plurality of PPSs. In that case, one of a plurality of PPSs is selected from each picture in the target sequence.

(Encoded picture)
In the coded picture, a set of data referred to by the image decoding device 31 in order to decode the picture PICT to be processed is defined. As shown in FIG. 1B, the picture PICT includes slices S0 to S _NS-1 (NS is the total number of slices included in the picture PICT).

In the following description, if it is not necessary to distinguish each of the slices S0 to _SNS-1 , the subscripts may be omitted. The same applies to data included in an encoded stream Te described below and to which other subscripts are attached.

(Encoded slice)
In the coded slice, a set of data referred to by the image decoding device 31 for decoding the slice S to be processed is defined. As shown in FIG. 1C, the slice S includes a slice header SH and slice data SDATA.

The slice header SH includes an encoding parameter group that is referred to by the image decoding device 31 in order to determine a decoding method of the target slice. Slice type designation information (slice_type) for designating a slice type is an example of an encoding parameter included in the slice header SH.

As slice types that can be specified by the slice type specification information, (1) I slice using only intra prediction at the time of encoding, (2) P slice using unidirectional prediction or intra prediction at the time of encoding, (3) B-slice using unidirectional prediction, bidirectional prediction, or intra prediction at the time of encoding may be used.

Note that the slice header SH may include a reference (pic_parameter_set_id) to the picture parameter set PPS included in the encoded video sequence.

(Encoded slice data)
In the encoded slice data, a set of data referred to by the image decoding device 31 for decoding the slice data SDATA to be processed is defined. The slice data SDATA includes a coding tree unit (CTU) as shown in FIG. A CTU is a block of a fixed size (for example, 64x64) that constitutes a slice, and is sometimes called a maximum coding unit (LCU: Large Coding Unit).

(Encoding tree unit)
As shown in (e) of FIG. 1, a set of data referred to by the image decoding device 31 in order to decode the encoding tree unit to be processed is defined. The coding tree unit is divided by recursive quadtree division. A tree-structured node obtained by recursive quadtree partitioning is referred to as a coding node (CN). An intermediate node of the quadtree is an encoding node, and the encoding tree unit itself is defined as the highest encoding node. The CTU includes a split flag (cu_split_flag), and when cu_split_flag is 1, it is split into four coding nodes CN. When cu_split_flag is 0, the coding node CN is not divided and has one coding unit (CU: Coding Unit) as a node. The encoding unit CU is a terminal node of the encoding node and is not further divided. The encoding unit CU is a basic unit of the encoding process.

In addition, when the size of the coding tree unit CTU is 64 × 64 pixels, the size of the coding unit can be any of 64 × 64 pixels, 32 × 32 pixels, 16 × 16 pixels, and 8 × 8 pixels.

(Encoding unit)
As shown in (f) of FIG. 1, a set of data referred to by the image decoding device 31 in order to decode an encoding unit to be processed is defined. Specifically, the encoding unit includes a prediction tree, a conversion tree, and a CU header CUH. The CU header defines a prediction mode, a division method (PU division mode), and the like.

In the prediction tree, prediction information (a reference picture index, a motion vector, etc.) of each prediction unit (PU) obtained by dividing the coding unit into one or a plurality is defined. In other words, the prediction unit is one or a plurality of non-overlapping areas constituting the encoding unit. The prediction tree includes one or a plurality of prediction units obtained by the above-described division. Hereinafter, a prediction unit obtained by further dividing the prediction unit is referred to as a “sub-block”. The sub block is composed of a plurality of pixels. When the sizes of the prediction unit and the sub-block are equal, the number of sub-blocks in the prediction unit is one. If the prediction unit is larger than the size of the sub-block, the prediction unit is divided into sub-blocks. For example, when the prediction unit is 8 × 8 and the sub-block is 4 × 4, the prediction unit is divided into four sub-blocks that are divided into two horizontally and two vertically.

The prediction process may be performed for each prediction unit (sub block).

There are roughly two types of division in the prediction tree: intra prediction and inter prediction. Intra prediction is prediction within the same picture, and inter prediction refers to prediction processing performed between different pictures (for example, between display times and between layer images).

In the case of intra prediction, there are 2Nx2N (the same size as the encoding unit) and NxN division methods.

Also, in the case of inter prediction, the division method is encoded by the PU division mode (part_mode) of encoded data, 2Nx2N (same size as the encoding unit), 2NxN, 2NxnU, 2NxnD, Nx2N, nLx2N, nRx2N, and NxN etc. 2NxN and Nx2N indicate 1: 1 symmetrical division,
2NxnU, 2NxnD and nLx2N, nRx2N show a 1: 3, 3: 1 asymmetric partitioning. The PUs included in the CU are expressed as PU0, PU1, PU2, and PU3 in this order.

(A) to (h) of FIG. 2 specifically illustrate the shape of the partition (the position of the boundary of the PU partition) in each PU partition mode. 2A shows a 2Nx2N partition, and FIGS. 2B, 2C, and 2D show 2NxN, 2NxnU, and 2NxnD partitions (horizontal partitions), respectively. (E), (f), and (g) show partitions (vertical partitions) in the case of Nx2N, nLx2N, and nRx2N, respectively, and (h) shows an NxN partition. The horizontal partition and the vertical partition are collectively referred to as a rectangular partition, and 2Nx2N and NxN are collectively referred to as a square partition.

Also, in the conversion tree, the encoding unit is divided into one or a plurality of conversion units, and the position and size of each conversion unit are defined. In other words, a transform unit is one or more non-overlapping areas that make up a coding unit. The conversion tree includes one or a plurality of conversion units obtained by the above division.

The division in the conversion tree includes a case where an area having the same size as that of the encoding unit is assigned as a conversion unit, and a case where recursive quadtree division is used, as in the case of the CU division described above.

Conversion processing is performed for each conversion unit.

(Prediction parameter)
A prediction image of a prediction unit (PU: Prediction Unit) is derived from a prediction parameter associated with the PU. The prediction parameters include a prediction parameter for intra prediction or a prediction parameter for inter prediction. Hereinafter, prediction parameters for inter prediction (inter prediction parameters) will be described. The inter prediction parameter includes prediction list use flags predFlagL0 and predFlagL1, reference picture indexes refIdxL0 and refIdxL1, and motion vectors mvL0 and mvL1. The prediction list use flags predFlagL0 and predFlagL1 are flags indicating whether or not reference picture lists called L0 list and L1 list are used, respectively, and a reference picture list corresponding to a value of 1 is used. In this specification, when “flag indicating whether or not it is XX” is described, when the flag is not 0 (for example, 1) is XX, 0 is not XX, and logical negation, logical product, etc. 1 is treated as true and 0 is treated as false (the same applies hereinafter). However, other values can be used as true values and false values in an actual apparatus or method.

Syntax elements for deriving inter prediction parameters included in the encoded data include, for example, PU partition mode part_mode, merge flag merge_flag, merge index merge_idx, inter prediction identifier inter_pred_idc, reference picture index refIdxLX, prediction vector index mvp_LX_idx, There is a difference vector mvdLX.

(Reference picture list)
The reference picture list is a list including reference pictures stored in the reference picture memory 306. FIG. 3 is a conceptual diagram illustrating an example of a reference picture and a reference picture list. In FIG. 3A, a rectangle is a picture, an arrow is a picture reference relationship, a horizontal axis is time, I, P, and B in the rectangle are intra pictures, uni-predictive pictures, bi-predictive pictures, and numbers in the rectangles are decoded. Indicates the order. As shown in the figure, the decoding order of pictures is I0, P1, B2, B3, and B4, and the display order is I0, B3, B2, B4, and P1. FIG. 3B shows an example of the reference picture list. The reference picture list is a list representing candidate reference pictures, and one picture (slice) may have one or more reference picture lists. In the illustrated example, the target picture B3 has two reference picture lists, an L0 list RefPicList0 and an L1 list RefPicList1. When the target picture is B3, the reference pictures are I0, P1, and B2, and the reference picture has these pictures as elements. In each prediction unit, which picture in the reference picture list RefPicListX is actually referred to is specified by the reference picture index refIdxLX. The figure shows an example in which reference pictures P1 and B2 are referenced by refIdxL0 and refIdxL1.

(Merge prediction and AMVP prediction)
The prediction parameter decoding (encoding) method includes a merge prediction (merge) mode and an AMVP (Adaptive Motion Vector Prediction) mode. The merge flag merge_flag is a flag for identifying these. The merge prediction mode is a mode that uses prediction parameters of already processed neighboring PUs without including the prediction list use flag predFlagLX (or inter prediction identifier inter_pred_idc), reference picture index refIdxLX, and motion vector mvLX in the encoded data. The mode is a mode in which the inter prediction identifier inter_pred_idc, the reference picture index refIdxLX, and the motion vector mvLX are included in the encoded data. The motion vector mvLX is encoded as a prediction vector index mvp_LX_idx for identifying the prediction vector mvpLX and a difference vector mvdLX.

The inter prediction identifier inter_pred_idc is a value indicating the type and number of reference pictures, and takes one of PRED_L0, PRED_L1, and PRED_BI. PRED_L0 and PRED_L1 indicate that reference pictures managed by the reference picture lists of the L0 list and the L1 list are used, respectively, and that one reference picture is used (single prediction). PRED_BI indicates that two reference pictures are used (bi-prediction BiPred), and reference pictures managed by the L0 list and the L1 list are used. The prediction vector index mvp_LX_idx is an index indicating a prediction vector, and the reference picture index refIdxLX is an index indicating a reference picture managed in the reference picture list. Note that LX is a description method used when L0 prediction and L1 prediction are not distinguished from each other. By replacing LX with L0 and L1, parameters for the L0 list and parameters for the L1 list are distinguished.

The merge index merge_idx is an index that indicates whether one of the prediction parameter candidates (merge candidates) derived from the processed PU is used as the prediction parameter of the decoding target PU.

(Motion vector)
The motion vector mvLX indicates a shift amount between blocks on two different pictures. A prediction vector and a difference vector related to the motion vector mvLX are referred to as a prediction vector mvpLX and a difference vector mvdLX, respectively.

(Inter prediction identifier inter_pred_idc and prediction list use flag predFlagLX)
The relationship between the inter prediction identifier inter_pred_idc and the prediction list use flags predFlagL0 and predFlagL1 is as follows and can be converted into each other.

inter_pred_idc = (predFlagL1 << 1) + predFlagL0
predFlagL0 = inter_pred_idc & 1
predFlagL1 = inter_pred_idc >> 1
Note that a prediction list use flag or an inter prediction identifier may be used as the inter prediction parameter. Further, the determination using the prediction list use flag may be replaced with the determination using the inter prediction identifier. Conversely, the determination using the inter prediction identifier may be replaced with the determination using the prediction list use flag.

(Determination of bi-prediction biPred)
The flag biPred as to whether it is a bi-prediction BiPred can be derived depending on whether the two prediction list use flags are both 1. For example, it can be derived by the following formula.

biPred = (predFlagL0 == 1 && predFlagL1 == 1)
The flag biPred can also be derived depending on whether or not the inter prediction identifier is a value indicating that two prediction lists (reference pictures) are used. For example, it can be derived by the following formula.

biPred = (inter_pred_idc == PRED_BI)? 1: 0
The above formula can also be expressed by the following formula.

biPred = (inter_pred_idc == PRED_BI)
For example, a value of 3 can be used for PRED_BI.

(Configuration of image decoding device)
Next, the configuration of the image decoding device 31 according to the present embodiment will be described. FIG. 5 is a schematic diagram illustrating a configuration of the image decoding device 31 according to the present embodiment. The image decoding device 31 includes an entropy decoding unit 301, a prediction parameter decoding unit (prediction image decoding device) 302, a loop filter 305, a reference picture memory 306, a prediction parameter memory 307, a prediction image generation unit (prediction image generation device) 308, and inversely. A quantization / inverse transform unit 311 and an adder 312 are included.

The prediction parameter decoding unit 302 includes an inter prediction parameter decoding unit 303 and an intra prediction parameter decoding unit 304. The predicted image generation unit 308 includes an inter predicted image generation unit 309 and an intra predicted image generation unit 310.

The entropy decoding unit 301 performs entropy decoding on the coded stream Te input from the outside, and separates and decodes individual codes (syntax elements). The separated codes include prediction information for generating a prediction image and residual information for generating a difference image.

The entropy decoding unit 301 outputs a part of the separated code to the prediction parameter decoding unit 302. Some of the separated codes are, for example, a prediction mode predMode, a PU partition mode part_mode, a merge flag merge_flag, a merge index merge_idx, an inter prediction identifier inter_pred_idc, a reference picture index refIdxLX, a prediction vector index mvp_LX_idx, and a difference vector mvdLX. Control of which code is decoded is performed based on an instruction from the prediction parameter decoding unit 302. The entropy decoding unit 301 outputs the quantized coefficient to the inverse quantization / inverse transform unit 311. This quantized coefficient is a coefficient obtained by performing transform (Discrete Cosine Transform, discrete cosine transform) on the residual signal and quantizing in the encoding process.

The inter prediction parameter decoding unit 303 decodes the inter prediction parameter with reference to the prediction parameter stored in the prediction parameter memory 307 based on the code input from the entropy decoding unit 301.

The inter prediction parameter decoding unit 303 outputs the decoded inter prediction parameter to the prediction image generation unit 308 and stores it in the prediction parameter memory 307. Details of the inter prediction parameter decoding unit 303 will be described later.

The intra prediction parameter decoding unit 304 refers to the prediction parameter stored in the prediction parameter memory 307 on the basis of the code input from the entropy decoding unit 301 and decodes the intra prediction parameter. The intra prediction parameter is a parameter used in a process of predicting a CU within one picture, for example, an intra prediction mode IntraPredMode. The intra prediction parameter decoding unit 304 outputs the decoded intra prediction parameter to the prediction image generation unit 308 and stores it in the prediction parameter memory 307.

The intra prediction parameter decoding unit 304 may derive different intra prediction modes depending on luminance and color difference. In this case, the intra prediction parameter decoding unit 304 decodes the luminance prediction mode IntraPredModeY as the luminance prediction parameter and the color difference prediction mode IntraPredModeC as the color difference prediction parameter. The luminance prediction mode IntraPredModeY is a 35 mode, and corresponds to planar prediction (0), DC prediction (1), and direction prediction (2 to 34). The color difference prediction mode IntraPredModeC uses one of the planar prediction (0), the DC prediction (1), the direction prediction (2 to 34), and the LM mode (35). The intra prediction parameter decoding unit 304 decodes a flag indicating whether IntraPredModeC is the same mode as the luminance mode. If the flag indicates that the mode is the same as the luminance mode, IntraPredModeC is assigned to IntraPredModeC, and the flag is luminance. If the mode is different from the mode, planar prediction (0), DC prediction (1), direction prediction (2 to 34), and LM mode (35) may be decoded as IntraPredModeC.

The loop filter 305 applies filters such as a deblocking filter, a sample adaptive offset (SAO), and an adaptive loop filter (ALF) to the decoded image of the CU generated by the adding unit 312.

The reference picture memory 306 stores the decoded image of the CU generated by the loop filter 305 at a predetermined position for each decoding target picture and CU.

The prediction parameter memory 307 stores the prediction parameter in a predetermined position for each decoding target picture and prediction unit (or sub-block, fixed-size block, pixel). Specifically, the prediction parameter memory 307 stores the inter prediction parameter decoded by the inter prediction parameter decoding unit 303, the intra prediction parameter decoded by the intra prediction parameter decoding unit 304, and the prediction mode predMode separated by the entropy decoding unit 301. . The stored inter prediction parameters include, for example, a prediction list utilization flag predFlagLX (inter prediction identifier inter_pred_idc), a reference picture index refIdxLX, and a motion vector mvLX.

The prediction image generation unit 308 receives the prediction mode predMode input from the entropy decoding unit 301 and the prediction parameter from the prediction parameter decoding unit 302. Further, the predicted image generation unit 308 reads a reference picture from the reference picture memory 306. The prediction image generation unit 308 generates a prediction image of the PU using the input prediction parameter and the read reference picture in the prediction mode indicated by the prediction mode predMode.

Here, when the prediction mode predMode indicates the inter prediction mode, the inter prediction image generation unit 309 uses the inter prediction parameter input from the inter prediction parameter decoding unit 303 and the read reference picture to perform prediction of the PU by inter prediction. Is generated.

The inter prediction image generation unit 309 performs a motion vector on the basis of the decoding target PU from the reference picture indicated by the reference picture index refIdxLX for a reference picture list (L0 list or L1 list) having a prediction list use flag predFlagLX of 1. The reference picture block at the position indicated by mvLX is read from the reference picture memory 306. The inter prediction image generation unit 309 performs prediction based on the read reference picture block to generate a prediction image of the PU. The inter prediction image generation unit 309 outputs the generated prediction image of the PU to the addition unit 312.

When the prediction mode predMode indicates the intra prediction mode, the intra predicted image generation unit 310 performs intra prediction using the intra prediction parameter input from the intra prediction parameter decoding unit 304 and the read reference picture. Specifically, the intra predicted image generation unit 310 reads, from the reference picture memory 306, neighboring PUs that are pictures to be decoded and are in a predetermined range from the decoding target PUs among the PUs that have already been decoded. The predetermined range is, for example, one of the left, upper left, upper, and upper right adjacent PUs when the decoding target PU sequentially moves in the so-called raster scan order, and differs depending on the intra prediction mode. The raster scan order is an order in which each row is sequentially moved from the left end to the right end in each picture from the upper end to the lower end.

The intra predicted image generation unit 310 performs prediction in the prediction mode indicated by the intra prediction mode IntraPredMode for the read adjacent PU, and generates a predicted image of the PU. The intra predicted image generation unit 310 outputs the generated predicted image of the PU to the adding unit 312.

When the intra prediction parameter decoding unit 304 derives an intra prediction mode different in luminance and color difference, the intra prediction image generation unit 310 performs planar prediction (0), DC prediction (1), direction according to the luminance prediction mode IntraPredModeY. Prediction image of luminance PU is generated by any of prediction (2 to 34), and planar prediction (0), DC prediction (1), direction prediction (2 to 34), LM mode according to color difference prediction mode IntraPredModeC A predicted image of the color difference PU is generated by any of (35).

The inverse quantization / inverse transform unit 311 inversely quantizes the quantized coefficient input from the entropy decoding unit 301 to obtain a transform coefficient. The inverse quantization / inverse transform unit 311 performs inverse transform (Inverse Discrete Cosine Transform) on the obtained transform coefficient to calculate a residual signal. The inverse quantization / inverse transform unit 311 outputs the calculated residual signal to the adder 312.

The addition unit 312 adds the prediction image of the PU input from the inter prediction image generation unit 309 or the intra prediction image generation unit 310 and the residual signal input from the inverse quantization / inverse conversion unit 311 for each pixel, Generate a decoded PU image. The loop filter 305 stores the generated decoded image of the PU in the reference picture memory 306, and outputs a decoded image Td obtained by integrating the generated decoded image of the PU for each picture to the outside.

(Configuration of inter prediction parameter decoding unit)
Next, the configuration of the inter prediction parameter decoding unit 303 will be described.

FIG. 12 is a schematic diagram illustrating a configuration of the inter prediction parameter decoding unit 303 according to the present embodiment. The inter prediction parameter decoding unit 303 includes an inter prediction parameter decoding control unit 3031, an AMVP prediction parameter derivation unit 3032, an addition unit 3035, a merge prediction parameter derivation unit 3036, a sub-block prediction parameter derivation unit 3037, and a BTM processing unit 3038. Composed.

The inter prediction parameter decoding control unit 3031 instructs the entropy decoding unit 301 to decode a code (syntax element) related to inter prediction, and a code (syntax element) included in the encoded data, for example, PU partition mode part_mode , Merge flag merge_flag, merge index merge_idx, inter prediction identifier inter_pred_idc, reference picture index refIdxLX, prediction vector index mvp_LX_idx, and difference vector mvdLX are extracted.

The inter prediction parameter decoding control unit 3031 first extracts a merge flag merge_flag. When the inter prediction parameter decoding control unit 3031 expresses that a certain syntax element is to be extracted, it means that the entropy decoding unit 301 is instructed to decode a certain syntax element, and the corresponding syntax element is read from the encoded data. To do.

When the merge flag merge_flag is 0, that is, indicates the AMVP prediction mode, the inter prediction parameter decoding control unit 3031 uses the entropy decoding unit 301 to extract the AMVP prediction parameter from the encoded data. Examples of AMVP prediction parameters include an inter prediction identifier inter_pred_idc, a reference picture index refIdxLX, a prediction vector index mvp_LX_idx, and a difference vector mvdLX. The AMVP prediction parameter derivation unit 3032 derives a prediction vector mvpLX from the prediction vector index mvp_LX_idx. Details will be described later. The inter prediction parameter decoding control unit 3031 outputs the difference vector mvdLX to the addition unit 3035. The adding unit 3035 adds the prediction vector mvpLX and the difference vector mvdLX to derive a motion vector.

When the merge flag merge_flag is 1, that is, indicates the merge prediction mode, the inter prediction parameter decoding control unit 3031 extracts the merge index merge_idx as a prediction parameter related to merge prediction. The inter prediction parameter decoding control unit 3031 outputs the extracted merge index merge_idx to the merge prediction parameter derivation unit 3036 (details will be described later), and outputs the sub-block prediction mode flag subPbMotionFlag to the sub-block prediction parameter derivation unit 3037. The subblock prediction parameter deriving unit 3037 divides the PU into a plurality of subblocks according to the value of the subblock prediction mode flag subPbMotionFlag, and derives a motion vector in units of subblocks. That is, in the sub-block prediction mode, the prediction block is predicted in units of blocks as small as 4x4 or 8x8. In the image encoding device 11 to be described later, a sub-block prediction mode is used for a method in which a CU is divided into a plurality of partitions (PUs such as 2NxN, Nx2N, and NxN) and the syntax of prediction parameters is encoded in units of partitions. Since a plurality of sub-blocks are grouped into a set and the syntax of the prediction parameter is encoded for each set, motion information of a large number of sub-blocks can be encoded with a small amount of code.

More specifically, the sub-block prediction parameter derivation unit 3037 performs at least one of a spatio-temporal sub-block prediction unit 30371, an affine prediction unit 30372, and a matching motion derivation unit 30373 that perform sub-block prediction in the sub-block prediction mode. Prepare.

(Subblock prediction mode flag)
Here, a method for deriving a sub-block prediction mode flag subPbMotionFlag indicating whether or not a prediction mode of a certain PU is a sub-block prediction mode in the image encoding device 11 (details will be described later) will be described. The image encoding device 11 derives a sub-block prediction mode flag subPbMotionFlag based on which of later-described spatial sub-block prediction SSUB, temporal sub-block prediction TSUB, affine prediction AFFINE, and matching motion derivation MAT is used. For example, when the prediction mode selected by a certain PU is N (for example, N is a label indicating the selected merge candidate), the sub-block prediction mode flag subPbMotionFlag may be derived by the following equation.

subPbMotionFlag = (N == TSUB) || (N == SSUB) || (N == AFFINE) || (N == MAT)
Here, || represents a logical sum (the same applies hereinafter).

Further, as described below, the above equation may be appropriately changed according to the type of sub-block prediction mode performed by the image encoding device 11. That is, when the image encoding device 11 is configured to perform spatial subblock prediction SSUB and affine prediction AFFINE, the subblock prediction mode flag subPbMotionFlag may be derived as follows.

subPbMotionFlag = (N == SSUB) || (N == AFFINE)
In addition, when prediction is performed by each prediction mode (for example, spatio-temporal sub-block prediction, affine prediction, matching motion derivation) included in the sub-block prediction, in the prediction mode processing corresponding to each sub-block prediction, The subPbMotionFlag may be set to 1.

Also, for example, when the CU size is 8x8 (logarithmic CU size log2CbSize == 3) and the PU is a small size such as a division type other than 2Nx2N, the PU can be a sub-block with a division number of 1. . In this case, the sub-block prediction mode subPbMotionFlag may be derived as follows.

subPbMotionFlag | = (log2CbSize == 3 && PartMode! = 2Nx2N)
Note that | = means that subPbMotionFlag may be derived by a sum operation (OR) with another condition. That is, subPbMotionFlag may be derived by the sum operation of determination of prediction mode N and small PU size determination as follows (the same applies hereinafter).

subPbMotionFlag = (N == TSUB) || (N == SSUB) || (N == AFFINE) || (N == MAT) || (log2CbSize == 3 && PartMode! = 2Nx2N)
Further, for example, a case where the CU size is 8 × 8 (log2CbSize == 3) and the division type is 2NxN, Nx2N, or NxN may be included in the sub-block prediction. That is, subPbMotionFlag may be derived as follows.

subPbMotionFlag | = (log2CbSize == 3 && (PartMode == 2NxN || PartMode == Nx2N || PartMode == NxN))
Further, for example, a case where the CU size is 8x8 (log2CbSize == 3) and the division type is NxN may be included in the sub-block prediction. That is, subPbMotionFlag may be derived as follows.

subPbMotionFlag | = (log2CbSize == 3 && PartMode == NxN)
In addition, a case where the PU width or height is 4 may be included as a case of determining sub-block prediction. That is, the sub-block prediction mode flag subPbMotionFlag may be derived as follows.

subPbMotionFlag | = (nPbW == 4 || nPbH == 4)
The sub-block prediction parameter deriving unit 3037 of the image decoding device 31 derives a sub-block prediction mode from the subPbMotionFlag by a method reverse to that described above.

(Sub-block prediction unit)
Next, the sub-block prediction unit will be described.

(Spatio-temporal sub-block prediction unit 30371)
The spatio-temporal sub-block prediction unit 30371 calculates the target PU from the motion vector of the PU on the reference image temporally adjacent to the target PU (for example, the immediately preceding picture) or the motion vector of the PU spatially adjacent to the target PU. The motion vector of the sub-block obtained by dividing is derived. Specifically, the motion vector spMvLX [xi] [yi] (xi = xPb) of each sub-block in the target PU is scaled by matching the motion vector of the PU on the reference image with the reference picture referenced by the target PU. + nSbW * i, yj = yPb + nSbH * j, i = 0, 1, 2, ..., nPbW / nSbW-1, j = 0, 1, 2, ..., nPbH / nSbH-1) Derived (temporal sub-block prediction). Here, (xPb, yPb) is the upper left coordinate of the target PU, nPbW, nPbH are the size of the target PU, and nSbW, nSbH are the sizes of the sub-blocks.

Also, by calculating a weighted average according to the distance between the motion vector of the PU adjacent to the target PU and the sub-block obtained by dividing the target PU, the motion vector spMvLX [ xi] [yi] (xi = xPb + nSbW * i, yj = yPb + nSbH * j, i = 0, 1, 2, ..., nPbW / nSbW-1, j = 0, 1, 2, ...・, NPbH / nSbH-1) may be derived (spatial sub-block prediction).

The above-described temporal sub-block prediction candidate TSUB and spatial sub-block prediction candidate SSUB are selected as one mode (merge candidate) of the merge mode.

(Affine prediction unit)
The affine prediction unit 30372 derives the affine prediction parameters of the target PU. In the present embodiment, motion vectors (mv0_x, mv0_y) (mv1_x, mv1_y) of two control points (V0, V1) of the target PU are derived as affine prediction parameters. Specifically, the motion vector of each control point may be derived by prediction from the motion vector of the adjacent PU of the target PU, and further, the prediction vector and the encoded data derived as the motion vector of the control point The motion vector of each control point may be derived from the sum of the difference vectors derived from.

FIG. 13 shows an example of deriving the motion vector spMvLX of each sub-block constituting the target PU (nPbW × nPbH) from the motion vector (mv0_x, mv0_y) of the control point V0 and the motion vector (mv1_x, mv1_y) of V1. FIG. The motion vector spMvLX of each subblock is derived as a motion vector for each point located at the center of each subblock, as shown in FIG.

The affine prediction unit 30372, based on the affine prediction parameters of the target PU, the motion vector spMvLX [xi] [yi] (xi = xPb + nSbW * i, yj = yPb + nSbH * j, i in the target PU = 0, 1, 2, ..., nPbW / nSbW-1, j = 0, 1, 2, ..., nPbH / nSbH-1) are derived using the following equations.

spMvLX [xi] [yi] [0] = mv0_x + (mv1_x-mv0_x) / nPbW * (xi + nSbW / 2)-(mv1_y-mv0_y) / nPbH * (yi + nSbH / 2)
spMvLX [xi] [yi] [1] = mv0_y + (mv1_y-mv0_y) / nPbW * (xi + nSbW / 2) + (mv1_x-mv0_x) / nPbH * (yi + nSbH / 2)
Here, xPb and yPb are the upper left coordinates of the target PU, nPbW and nPbH are the width and height of the target PU, and nSbW and nSbH are the width and height of the sub-block.

(Matching motion deriving unit 30373)
The matching motion deriving unit 30373 derives a motion vector spMvLX of the sub-blocks constituting the PU by performing either bilateral matching or template matching. FIG. 14 is a diagram for explaining (a) bilateral matching and (b) template matching. The matching motion derivation mode is selected as one merge candidate (matching candidate) in the merge mode.

The matching motion deriving unit 30373 derives a motion vector by matching regions in a plurality of reference images, assuming that the object moves at a constant velocity. In bilateral matching, it is assumed that a certain object passes through a certain region of the reference image A, a target PU of the target picture Cur_Pic, and a certain region of the reference image B with constant velocity motion, and matching between the reference images A and B To derive the motion vector of the target PU. In the template matching, assuming that the adjacent vector of the target PU and the motion vector of the target PU are equal, a motion vector is derived by matching the adjacent region Temp_Cur of the target PU with the adjacent region Temp_L0 of the reference block on the reference picture. The matching motion deriving unit 30373 divides the target PU into a plurality of sub-blocks, and performs bilateral matching or template matching (to be described later) in units of the divided sub-blocks, thereby sub-block motion vectors spMvLX [xi] [yi] ( xi = xPb + nSbW * i, yj = yPb + nSbH * j, i = 0, 1, 2, ..., nPbW / nSbW-1, j = 0, 1, 2, ..., nPbH / nSbH- Derive 1).

As shown in FIG. 14A, in bilateral matching, two reference images are referred to in order to derive a motion vector of the sub-block Cur_block in the current picture Cur_Pic. More specifically, first, when the coordinates of the sub-block Cur_block are expressed as (xCur, yCur), an area in the reference image Ref0 (referred to as reference picture A) specified by the reference picture index refIdxL0,
(XPos0, yPos0) = (xCur + MV0_x, yCur + MV0_y)
Block_A having the upper left coordinates (xPos0, yPos0) specified by, and a reference image Ref1 (referred to as reference picture B) specified by the reference picture index refIdxL1,
(XPos1, yPos1) = (xCur + MV1_x, xCur + MV1_y) = (xCur-MV0_x * TD1 / TD0, yCur-MV0_y * TD1 / TD0)
Block_B having the upper left coordinates (xPos1, yPos1) specified by is set. Here, TD0 and TD1 represent the inter-picture distance between the target picture Cur_Pic and the reference picture A, and the inter-picture distance between the target picture Cur_Pic and the reference picture B, respectively, as shown in FIG. ing.

Next, (MV0_x, MV0_y) is determined so that the matching cost between Block_A and Block_B is minimized. The (MV0_x, MV0_y) derived in this way is a motion vector assigned to the sub-block.

On the other hand, (b) of FIG. 14 is a figure for demonstrating a template matching (Template | matching) among the said matching processes.

As shown in FIG. 14B, in template matching, one reference picture is referred to at a time in order to derive a motion vector of the sub-block Cur_block in the target picture Cur_Pic. That is, processing is performed for each reference picture.

More specifically, first, an area in a reference image Ref0 (referred to as reference picture A) designated by a reference picture index refIdxL0,
(XPos, yPos) = (xCur + MV0_x, yCur + MV0_y)
The reference block Block_A having the upper left coordinates (xPos, yPos) specified by is specified. Here, (xCur, yCur) is the upper left coordinate of the sub-block Cur_block.

Next, a template region Temp_Cur adjacent to the sub-block Cur_block in the target picture Cur_Pic and a template region Temp_L0 adjacent to the Block_A in the reference picture A are set. In the example shown in FIG. 14B, the template region Temp_Cur is composed of a region adjacent to the upper side of the sub-block Cur_block and a region adjacent to the left side of the sub-block Cur_block. The template area Temp_L0 is composed of an area adjacent to the upper side of Block_A and an area adjacent to the left side of Block_A.

Next, it is determined that the matching cost between Temp_Cur and TempL0 is minimum (MV0_x, MV0_y), and the motion vector is given to the sub-block spMvL0.

In template matching, the matching motion deriving unit 30373 may perform the same processing as the reference image Ref0 on the reference image Ref1 different from the reference image Ref0. In this case, an area in the reference image Ref1 (referred to as reference picture A) designated by the reference picture index refIdxL1,
(XPos, yPos) = (xCur + MV1_x, yCur + MV1_y)
The reference block Block_A having the upper left coordinates (xPos, yPos) specified by is specified. Then, a template area Temp_L1 adjacent to Block_A in the reference picture A is set.

Finally, (MV1_x, MV1_y) that minimizes the matching cost between Temp_Cur and TempL1 is determined, and (MV1_x, MV1_y) becomes the motion vector spMvL1 assigned to the sub-block.

Also, the matching motion derivation unit 30373 may process the two reference images Ref0 and Ref1 in the template matching. In this case, the matching motion deriving unit 30373 sequentially performs the matching of one reference image Ref0 and the matching of one reference image Ref1 described above.

FIG. 7 is a schematic diagram illustrating the configuration of the merge prediction parameter deriving unit 3036 according to the present embodiment. The merge prediction parameter derivation unit 3036 includes a merge candidate derivation unit 30361, a merge candidate selection unit 30362, and a merge candidate storage unit 30363. The merge candidate storage unit 30363 stores the merge candidates input from the merge candidate derivation unit 30361. The merge candidate includes a prediction list use flag predFlagLX, a motion vector mvLX, and a reference picture index refIdxLX. In the merge candidate storage unit 30363, an index is assigned to the stored merge candidate according to a predetermined rule.

The merge candidate derivation unit 30361 derives a merge candidate using the motion vector of the adjacent PU that has already been decoded and the reference picture index refIdxLX as they are. In addition, merge candidates may be derived using affine prediction. This method is described in detail below. The merge candidate derivation unit 30361 may use affine prediction for a spatial merge candidate derivation process, a temporal merge candidate derivation process, a combined merge candidate derivation process, and a zero merge candidate derivation process described later. Affine prediction is performed in units of sub-blocks, and prediction parameters are stored in the prediction parameter memory 307 for each sub-block. Alternatively, the affine prediction may be performed on a pixel basis.

(Spatial merge candidate derivation process)
As the spatial merge candidate derivation process, the merge candidate derivation unit 30361 reads and reads the prediction parameters (prediction list use flag predFlagLX, motion vector mvLX, reference picture index refIdxLX) stored in the prediction parameter memory 307 according to a predetermined rule. The predicted parameters are derived as merge candidates. The prediction parameter to be read is a prediction parameter related to each of the PUs within a predetermined range from the decoding target PU (for example, all or part of the PUs in contact with the lower left end, the upper left end, and the upper right end of the decoding target PU, respectively). is there. The merge candidates derived by the merge candidate deriving unit 30361 are stored in the merge candidate storage unit 30363.

(Time merge candidate derivation process)
As the temporal merge derivation process, the merge candidate derivation unit 30361 reads the prediction parameter of the PU in the reference image including the lower right coordinate of the decoding target PU from the prediction parameter memory 307 and sets it as a merge candidate. The reference picture designation method may be, for example, the reference picture index refIdxLX designated in the slice header, or may be designated using the smallest reference picture index refIdxLX of the PU adjacent to the decoding target PU. The merge candidates derived by the merge candidate deriving unit 30361 are stored in the merge candidate storage unit 30363.

(Join merge candidate derivation process)
As the merge merge derivation process, the merge candidate derivation unit 30361 uses two different derived merge candidate motion vectors and reference picture indexes already derived and stored in the merge candidate storage unit 30363 as the motion vectors of L0 and L1, respectively. Combined merge candidates are derived by combining them. The merge candidates derived by the merge candidate deriving unit 30361 are stored in the merge candidate storage unit 30363.

(Zero merge candidate derivation process)
As the zero merge candidate derivation process, the merge candidate derivation unit 30361 derives a merge candidate in which the reference picture index refIdxLX is 0 and both the X component and the Y component of the motion vector mvLX are 0. The merge candidates derived by the merge candidate deriving unit 30361 are stored in the merge candidate storage unit 30363.

The merge candidate selection unit 30362 selects, from the merge candidates stored in the merge candidate storage unit 30363, a merge candidate to which an index corresponding to the merge index merge_idx input from the inter prediction parameter decoding control unit 3031 is assigned. As an inter prediction parameter. The merge candidate selection unit 30362 stores the selected merge candidate in the prediction parameter memory 307 and outputs it to the prediction image generation unit 308.

FIG. 8 is a schematic diagram showing the configuration of the AMVP prediction parameter derivation unit 3032 according to this embodiment. The AMVP prediction parameter derivation unit 3032 includes a vector candidate derivation unit 3033, a vector candidate selection unit 3034, and a vector candidate storage unit 3039. The vector candidate derivation unit 3033 reads the already processed PU motion vector mvLX stored in the prediction parameter memory 307 based on the reference picture index refIdx, derives a prediction vector candidate, and sends the prediction vector candidate to the vector candidate storage unit 3039. Store in candidate list mvpListLX [].

The vector candidate selection unit 3034 selects the motion vector mvpListLX [mvp_LX_idx] indicated by the prediction vector index mvp_LX_idx from the prediction vector candidates in the prediction vector candidate list mvpListLX [] as the prediction vector mvpLX. The vector candidate selection unit 3034 outputs the selected prediction vector mvpLX to the addition unit 3035.

Note that a prediction vector candidate is a PU for which decoding processing has been completed, and is derived by scaling a motion vector of a PU (for example, an adjacent PU) within a predetermined range from the decoding target PU. The adjacent PU includes a PU that is spatially adjacent to the decoding target PU, for example, the left PU and the upper PU, and an area that is temporally adjacent to the decoding target PU, for example, the same position as the decoding target PU. It includes areas obtained from prediction parameters of PUs with different times.

The addition unit 3035 adds the prediction vector mvpLX input from the AMVP prediction parameter derivation unit 3032 and the difference vector mvdLX input from the inter prediction parameter decoding control unit 3031 to calculate a motion vector mvLX. The adding unit 3035 outputs the calculated motion vector mvLX to the predicted image generation unit 308 and the prediction parameter memory 307.

Note that the motion vector derived by the merge prediction parameter deriving unit 3036 or the motion vector derived by the matching motion deriving unit 30373 is not directly output to the inter-predicted image generation unit 309 but is output via the BTM processing unit 3038. May be.

The BTM processing unit 3038 uses the predicted image generated using the motion vector derived by the merge prediction parameter deriving unit 3036 or the matching motion deriving unit 30373 as a template, and executes bilateral template matching (BTM) processing. A highly accurate motion vector is derived. Details of the BTM processing will be described later. BTM is also called Decoder-side motion vector refinement (DMVR).

(Inter prediction image generation unit 309)
FIG. 11 is a schematic diagram illustrating a configuration of an inter predicted image generation unit 309 included in the predicted image generation unit 308 according to the present embodiment. The inter prediction image generation unit 309 includes a motion compensation unit (prediction image generation device) 3091 and a weight prediction unit 3094.

(Motion compensation)
The motion compensation unit 3091 receives the reference picture index refIdxLX from the reference picture memory 306 based on the inter prediction parameters (prediction list use flag predFlagLX, reference picture index refIdxLX, motion vector mvLX) input from the inter prediction parameter decoding unit 303. In the reference picture specified in (2), an interpolation image (motion compensated image, predicted image candidate) predSamplesLX is generated by reading out a block at a position shifted by the motion vector mvLX starting from the position of the decoding target PU. Here, when the accuracy of the motion vector mvLX is not integer accuracy, a motion compensation image is generated by applying a filter for generating a pixel at a decimal position called a motion compensation filter.

(Weight prediction)
The weight prediction unit 3094 generates a predicted image of the PU by multiplying the input motion compensated image (interpolated image, predicted image candidate) predSamplesLX by a weight coefficient. When one of the prediction list use flags (predFlagL0 or predFlagL1) is 1 (in the case of single prediction) and the weight prediction is not used, the input motion compensated image predSamplesLX (LX is L0 or L1) is represented by the number of pixel bits bitDepth The following equation is processed to match

predSamples [x] [y] = Clip3 (0, (1 << bitDepth)-1, (predSamplesLX [x] [y] + offset1) >> shift1)
Here, shift1 = 14−bitDepth, offset1 = 1 << (shift1-1).
When both of the reference list use flags (predFlagL0 and predFlagL1) are 1 (in the case of bi-prediction BiPred) and weight prediction is not used, the input motion compensation images predSamplesL0 and predSamplesL1 are averaged and the number of pixel bits The following equation is processed to match

predSamples [x] [y] = Clip3 (0, (1 << bitDepth)-1, (predSamplesL0 [x] [y] + predSamplesL1 [x] [y] + offset2) >> shift2)
Here, shift2 = 15-bitDepth, offset2 = 1 << (shift2-1).

Furthermore, in the case of single prediction, when performing weight prediction, the weight prediction unit 3094 derives the weight prediction coefficient w0 and the offset o0 from the encoded data, and performs the processing of the following equation.

predSamples [x] [y] = Clip3 (0, (1 << bitDepth)-1, ((predSamplesLX [x] [y] * w0 + 2 ^ (log2WD-1)) >> log2WD) + o0)
Here, log2WD is a variable indicating a predetermined shift amount.

Furthermore, in the case of bi-prediction BiPred, when performing weight prediction, the weight prediction unit 3094 derives weight prediction coefficients w0, w1, o0, o1 from the encoded data, and performs the processing of the following equation.

predSamples [x] [y] = Clip3 (0, (1 << bitDepth)-1, (predSamplesL0 [x] [y] * w0 + predSamplesL1 [x] [y] * w1 + ((o0 + o1 + 1) << log2WD)) >> (log2WD + 1))
<Motion vector decoding process>
Below, with reference to FIG. 9, the motion vector decoding process which concerns on this embodiment is demonstrated concretely.

As is clear from the above description, the motion vector decoding process according to the present embodiment includes a process of decoding syntax elements related to inter prediction (also referred to as motion syntax decoding process) and a process of deriving a motion vector ( Motion vector derivation process).

(Motion syntax decoding process)
FIG. 9 is a flowchart illustrating a flow of inter prediction syntax decoding processing performed by the inter prediction parameter decoding control unit 3031. In the following description of FIG. 9, each process is performed by the inter prediction parameter decoding control unit 3031 unless otherwise specified.

First, in step S101, the merge flag merge_flag is decoded, and in step S102,
merge_flag! = 0 (whether merge_flag is not 0)
Is judged.

When merge_flag! = 0 is true (Y in S102), the merge index merge_idx is decoded in S103, and the motion vector derivation process (S111) in the merge mode is executed.

When merge_flag! = 0 is false (N in S102), the inter prediction identifier inter_pred_idc is decoded in S104.

When inter_pred_idc is other than PRED_L1 (PRED_L0 or PRED_BI), the reference picture index refIdxL0, the difference vector parameter mvdL0, and the prediction vector index mvp_L0_idx are decoded in S105, S106, and S107, respectively.

When inter_pred_idc is other than PRED_L0 (PRED_L1 or PRED_BI), the reference picture index refIdxL1, the difference vector parameter mvdL1, and the prediction vector index mvp_L1_idx are decoded in S108, S109, and S110. Subsequently, a motion vector derivation process (S112) in the AMVP mode is executed.

(Example of motion vector derivation processing)
The motion vector deriving process can be derived as follows. FIG. 15 is a flowchart showing an outline of the motion prediction mode determination flow. The motion prediction mode determination flow is executed by the inter prediction parameter decoding unit 303. The motion prediction mode is a mode for determining a method for deriving a motion vector used for motion compensation prediction.

As shown in FIG. 15, in the motion prediction mode determination flow, first, the inter prediction parameter decoding control unit 3031 decodes the merge flag merge_flag (S1400), and determines whether the mode is the merge mode (S1401). Specifically, the inter prediction parameter decoding control unit 3031 decodes the merge flag merge_flag and determines merge_flag == 1, thereby determining whether or not the mode is the merge mode.

If it is determined that the mode is not the merge mode (NO in S1401), the AMVP mode is set. Specifically, when merge_flag == 1 is false, the AMVP mode (S1411) is selected. In this case, the inter prediction parameter decoding control unit 3031 decodes the difference vector mvdLX and decodes the prediction vector index mvp_LX_idx. Further, the AMVP prediction parameter deriving unit 3032 derives a prediction vector mvpLX and derives a motion vector mvLX by the following equation.

mvLX = pmvCand [mvp_LX_idx] + mvdLX
Note that mvpLX = pmvCand [mvp_LX_idx].

On the other hand, if it is determined that the mode is the merge mode (YES in S1401), the inter prediction parameter decoding control unit 3031 decodes the FLUC (FRUC: Frame Rate Up Conversion) mode index (S1402), and is the matching motion derivation mode? It is determined whether or not (S1403). Specifically, when merge_flag == 1 is true, the flack mode index fruc_mode_idx is decoded, and fruc_mode_idx! = 0 is determined.

If it is determined that the mode is the matching motion derivation mode (YES in S1403), the matching motion derivation mode is set (S1404). Specifically, when fruc_mode_idx! = 0 is true, the matching motion derivation mode is selected as the motion vector derivation method. Then, the matching motion deriving unit 30373 determines whether to perform template matching (TM) (S1404a). When performing template matching (YES in S1404a), the matching motion deriving unit 30373 performs template matching (S1404b), and when not performing template matching (NO in S1404a), the matching motion deriving unit 30373 performs bilateral matching (BM). This is performed (S1404c). In step S1404, when the fruc_mode_idx is MODE_BM (for example, 1), the pattern matching vector is derived by bilateral matching when the fruc_mode_idx is MODE_BM (for example, 1). Is done.

In step 1403, if it is determined that the mode is not the matching motion derivation mode (NO in S1403), the merge mode is set (S1412). Specifically, when fruc_mode_idx! = 0 is false, the inter prediction parameter decoding control unit 3031 decodes the merge index merge_idx. Subsequently, the merge prediction parameter deriving unit 3036 derives merge candidates mergeCand and derives a motion vector mvLX by the following equation.

mvLX = mergeCand [merge_idx]
Then, regarding the motion vector derived in the merge mode, the BTM processing unit 3038 determines whether or not to perform bilateral template matching (BTM) processing (S1413), and determines that BTM processing is performed (YES in S1413). , BTM processing is executed (S1414).

Whether or not to perform BTM processing is determined to perform BTM processing if a motion vector is derived by bi-directional prediction in the merge mode, and is determined not to perform BTM processing unless bi-directional prediction is performed. . In addition, in the case of the prediction parameter for luminance compensation and the affine prediction parameter, the BTM process is not performed even in bidirectional prediction.

(BTM (bilateral template matching) processing)
Next, details of the bilateral template matching (BTM) process will be described with reference to FIGS. 16 and 17. FIG. 16A is a diagram showing a relationship between a reference image and a template in BTM processing, and FIG. 16B is a diagram showing a processing flow. FIG. 17 is a diagram for explaining the details of the template in the BTM processing.

As shown in FIG. 16A and FIG. 17, the BTM processing unit 3038 first generates a prediction block of the target block Cur ブロック block from a plurality of reference images Ref0 and Ref1 derived by the merge prediction parameter deriving unit 3036. This is used as a template.

Specifically, the processing is the same as the processing of the weight prediction unit 3094, the motion compensated image predSamplesL0 [] [] derived from the reference image Ref0 and the motion vector mvL0, and the motion derived from the reference image Ref1 and the motion vector mvL1. The weight prediction image predSamples [] [] derived from the compensation image predSamplesL1 [] [] is defined as a template predBTM [] [].
predBTM [x] [y] = Clip3 (0, (1 << bitDepth)-1, (predSamplesL0 [x] [y] + predSamplesL1 [x] [y] + offset2) >> shift2)
Referring to the above equation, the coefficients of predSamplesL0 [x] [y] and predSamplesL1 [x] [y] are 1. That is, it can be said that the template predBTM [x] [y] is derived from the average value of predSamplesL0 [x] [y] and predSamplesL1 [x] [y].

Next, the BTM processing unit 3038 sets motion vector candidates in a range centered on mvL0 or mvL1 (initial vector) for each of the reference image Ref0 and the reference image Ref1, Deriving the matching cost with the template. Then, the motion vectors mvL0 and mvL1 are updated to the vectors mvL0 ′ and mvL1 ′ that minimize the matching cost.

Next, the flow of BTM processing will be described with reference to FIG. As shown in FIG. 16B, the BTM processing unit 3038 first derives a template (S501). As described above, the template is a prediction block derived from a plurality of motion vectors (for example, mvL0 and mvL1) derived by the merge prediction parameter deriving unit 3036. Next, the BTM processing unit 3038 performs a local search (S502). As shown in FIG. 16B, the local search is performed in the order of the L0 search process (S503) for the reference image Ref0 and the L1 search process (S504) for the reference image Ref1. In the L0 search process and the L1 search process, an optimum vector bestMV that minimizes the matching cost is searched. That is, the BTM processing unit 3038 derives a matching cost between the template predBTM [] [] and the motion compensated image predSamplesLX [] [] for the motion vector candidate of the reference image RefX with X = 0..1, and the matching cost Let the motion vector candidate that minimizes be the optimal vector bestMV. As matching costs for the optimal vector, the matching cost for the reference image Ref0 is called costL0, and the matching cost for the reference image Ref1 is called costL1. Hereinafter, the matching cost for the reference image RefX is also expressed as costLX.

The inter prediction image generation unit 309 obtains the motion compensated image predSamplesL0 [x] [y] derived using the motion vector mvL0 obtained from the L0 search processing by the BTM processing unit 3038 and the L1 search processing by the BTM processing unit 3038. The final predicted image predSamples [x] [y] is derived from the motion compensated image predSamplesL1 [x] [y] derived using the motion vector mvL1 obtained. For the process of deriving the predicted image predSamples [x] [y] from the two motion compensated images predSamplesL1 [x] [y] and predSamplesL1 [x] [y], the method described above may be used. The method described may be used.

(TM (template matching) processing)
Next, an overview of template matching (TM) will be described with reference to FIG. 18A and 18B are diagrams showing an outline of template matching. As shown in FIG. 18B, in the template matching, the L0 matching process (S801) related to the reference image Ref0 and the L1 matching process (S802) related to the reference image Ref1 are sequentially performed. In each of the L0 matching process and the L1 matching process, a block search (S811, S821) and a sub-block search (S812, S822) are performed. In the L0 matching process and the L1 matching process, the matching motion derivation unit 30373 searches for the optimum vector bestMV that minimizes the matching cost. As in the case of BTM, the matching cost for the reference image Ref0 when it is an optimal vector is called costL0, and the matching cost for the reference image Ref1 is called costL1.

In this specification, the motion vector correction or derivation in the reference image RefX is called “LX search processing” and TM is “LX matching processing”, but both use a template for the reference image RefX. Matching process. That is, except for the fact that the matching area (template) and the initial vector are different between BTM and TM, these are almost the same processing.

As in the case of BTM, the inter prediction image generation unit 309 performs the motion compensated image predSamplesL0 [x] [y] based on the motion vector mvL0 obtained from the L0 matching process and the motion compensated image predSamplesL1 based on the motion vector mvL1 obtained from the L1 matching process. The final predicted image predSamples [x] [y] is derived from [x] [y]. For the process of deriving the predicted image predSamples [x] [y] from the two motion compensated images predSamplesL1 [x] [y] and predSamplesL1 [x] [y], the method described above may be used. The method described may be used.

(BM (bilateral matching) processing)
Next, an outline of bilateral matching (BM) will be described with reference to FIG. (A) and (b) of FIG. 19 are diagrams showing an outline of bilateral matching. As shown in FIG. 19B, in bilateral matching, L0 / L1 matching processing (S901) is performed. In the L0 / L1 matching process, a block search (S911) and a sub-block search (S912) are performed.

(Details of TM and BM)
Next, details of template matching and bilateral matching will be described with reference to FIGS. 20 and 21. FIG. FIG. 20 is a flowchart showing the flow of motion vector derivation processing in the matching motion derivation mode (template matching, bilateral matching). FIG. 21 is a diagram for explaining motion vector derivation processing in the matching motion derivation mode.

20 is executed by the matching motion deriving unit 30373.

Of steps shown in FIG. 20, steps S1051 to S1054 are block searches executed at the block level. That is, as shown in (a) of FIG. 21, a motion vector in a target block (CU or PU) unit is derived.

Also, steps S1055 to S1060 are a sub-block search executed at the sub-block level. That is, as shown in (b) of FIG. 21, a motion vector is derived in units of subblocks constituting the target block. Note that the size of the sub-block is 1/8 both vertically and horizontally with respect to the target block. However, the minimum size of the sub-block is 4 × 4 pixels. In FIG. 20, FRUC_MODE is a variable indicating the type of matching motion derivation mode, and corresponds to the fruc_mode_idx already described.

First, in step S1051, FRUC_MODE == MODE_TM is determined. If FRUC_MODE == MODE_TM is true (YES in S1051), a template for template matching is derived in step S1052. More specifically, a template for template matching is derived from the peripheral area of the block. Specifically, as shown in FIG. 21C, a template is derived from the upper adjacent area and the left adjacent area of the target block. Specifically, the pixels of the upper adjacent area and the left adjacent area of the target block are copied to form a template image. Note that the upper adjacent area and the left adjacent area of the target block are used only when their adjacency is available. When one of the areas cannot be used, a pixel image is copied from the upper adjacent area or the left adjacent area of the target block.

It should be noted that the pixel width (template size) copied from the adjacent area is suitably 4 pixels. Then, the process proceeds to step S1053.

Also, if FRUC_MODE == MODE_TM is false (NO in S1051), the process proceeds to step S1053. In step S1053, a block-level initial vector in the target block is derived (initial vector search). Note that the initial vector is a motion vector that serves as a base for the search. From a limited motion vector candidate (spatial merge candidate, temporal merge candidate, combined merge candidate, zero vector, ATMVP vector of the target block, etc.), a matching cost is determined. The vector that minimizes is derived as the initial vector. As described above, the initial vector candidate is a motion vector derived based on the motion vector of the processed reference point. That is, it is derived from motion vectors of already processed points (processed motion vectors) like various merge candidates, a scaled version of a processed motion vector, or a plurality of processed motion vectors such as ATMVP vectors. Use representative values. The ATMVP vector is a vector derived from the average (or weighted average, median) of the motion vector around the target block and the motion vector of the reference image. The initial vector search does not include a step search that repeats the search recursively according to the searched result and a raster search that searches a continuous region. These searches are called local searches.

In step S1054, a block level local search (local search) in the target block is performed. In the local search, a local region centered on the initial vector derived in step S1051 is further searched to search for a vector having a minimum matching cost, and set as a final vector. The local search may be a step search, raster search, or spiral search. Details of the local search will be described later.

Subsequently, the following processing is performed for each sub-block included in the target block (steps S1055 to S1060).

First, in step S1056, FRUC_MODE == MODE_TM is determined. If FRUC_MODE == MODE_TM is true (YES in S1056), a template for performing template matching is acquired in step S1057. More specifically, a template for template matching is acquired from the peripheral area of the sub-block. Specifically, as shown in FIG. 21C, a template is acquired from the upper adjacent region or the left adjacent region of the target sub block. Then, the process proceeds to step S1058. Also, if FRUC_MODE == MODE_TM is false (NO in S1056), the process proceeds to step S1058.

In step S1058, an initial vector of a sub-block in the target block is derived (initial vector search). In detail, vector candidates (target block selection vector, zero vector, center collocation vector of the subblock, lower right collocation vector of the subblock, ATMVP vector of the subblock, upper adjacent vector of the subblock, the sub The vector having the smallest matching cost among the left adjacent vectors of the block is set as the initial vector of the sub-block. Note that the vector candidates used for the initial vector search of the sub-block are not limited to the vectors described above.

Next, in step S1059, a local search centering on the initial vector of the sub-block selected in S1058 is performed. Then, the matching cost of vector candidates near the initial vector of the sub-block is derived, and the minimum vector is derived as the sub-block motion vector.

Then, when the process is completed for all the sub-blocks included in the target block, the pattern match vector derivation process ends.

Note that, for both initial vector search and local search, if FRUC_MODE == MODE_TM, the matching cost is derived by bilateral matching. If FRUC_MODE == MODE_TM, the matching cost is derived by template matching.

(Configuration of image encoding device)
Next, the configuration of the image encoding device 11 according to the present embodiment will be described. FIG. 4 is a block diagram illustrating a configuration of the image encoding device 11 according to the present embodiment. The image encoding device 11 includes a predicted image generation unit 101, a subtraction unit 102, a transform / quantization unit 103, an entropy encoding unit 104, an inverse quantization / inverse transform unit 105, an addition unit 106, a loop filter 107, and a prediction parameter memory. (Prediction parameter storage unit, frame memory) 108, reference picture memory (reference image storage unit, frame memory) 109, encoding parameter determination unit 110, and prediction parameter encoding unit 111. The prediction parameter encoding unit 111 includes an inter prediction parameter encoding unit 112 and an intra prediction parameter encoding unit 113.

The predicted image generation unit 101 generates, for each picture of the image T, a predicted image P of the prediction unit PU for each encoding unit CU that is an area obtained by dividing the picture. Here, the predicted image generation unit 101 reads a decoded block from the reference picture memory 109 based on the prediction parameter input from the prediction parameter encoding unit 111. The prediction parameter input from the prediction parameter encoding unit 111 is, for example, a motion vector in the case of inter prediction. The predicted image generation unit 101 reads a block at a position on the reference image indicated by the motion vector with the target PU as a starting point. In the case of intra prediction, the prediction parameter is, for example, an intra prediction mode. A pixel value of an adjacent PU used in the intra prediction mode is read from the reference picture memory 109, and a predicted image P of the PU is generated. The predicted image generation unit 101 generates a predicted image P of the PU using one prediction method among a plurality of prediction methods for the read reference picture block. The predicted image generation unit 101 outputs the generated predicted image P of the PU to the subtraction unit 102.

Note that the predicted image generation unit 101 performs the same operation as the predicted image generation unit 308 already described. For example, FIG. 6 is a schematic diagram illustrating a configuration of an inter predicted image generation unit 1011 included in the predicted image generation unit 101. The inter prediction image generation unit 1011 includes a motion compensation unit 10111 and a weight prediction unit 10112. Since the motion compensation unit 10111 and the weight prediction unit 10112 have the same configurations as the motion compensation unit 3091 and the weight prediction unit 3094 described above, description thereof is omitted here. Further, the configuration of the inter predicted image generation unit 1011 may include the same configuration as the inter predicted image generation unit 309 described in the second and third embodiments described later. In this case, the configuration of the inter prediction parameter encoding unit 112 may include the same configuration as that of the inter prediction parameter decoding unit 303.

The prediction image generation unit 101 generates a prediction image P of the PU based on the pixel value of the reference block read from the reference picture memory, using the parameter input from the prediction parameter encoding unit. The predicted image generated by the predicted image generation unit 101 is output to the subtraction unit 102 and the addition unit 106.

The subtraction unit 102 subtracts the signal value of the predicted image P of the PU input from the predicted image generation unit 101 from the pixel value of the corresponding PU of the image T, and generates a residual signal. The subtraction unit 102 outputs the generated residual signal to the transform / quantization unit 103.

The transformation / quantization unit 103 performs transformation on the residual signal input from the subtraction unit 102 and calculates a transformation coefficient. The transform / quantization unit 103 quantizes the calculated transform coefficient to obtain a quantized coefficient. The transform / quantization unit 103 outputs the obtained quantization coefficient to the entropy coding unit 104 and the inverse quantization / inverse transform unit 105.

The entropy encoding unit 104 receives the quantization coefficient from the transform / quantization unit 103 and receives the encoding parameter from the prediction parameter encoding unit 111. Examples of input encoding parameters include codes such as a reference picture index refIdxLX, a prediction vector index mvp_LX_idx, a difference vector mvdLX, a prediction mode predMode, and a merge index merge_idx.

The entropy encoding unit 104 generates an encoded stream Te by entropy encoding the input quantization coefficient and encoding parameter, and outputs the generated encoded stream Te to the outside.

The inverse quantization / inverse transform unit 105 inversely quantizes the quantization coefficient input from the transform / quantization unit 103 to obtain a transform coefficient. The inverse quantization / inverse transform unit 105 performs inverse transform on the obtained transform coefficient to calculate a residual signal. The inverse quantization / inverse transform unit 105 outputs the calculated residual signal to the addition unit 106.

The addition unit 106 adds the signal value of the prediction image P of the PU input from the prediction image generation unit 101 and the signal value of the residual signal input from the inverse quantization / inverse conversion unit 105 for each pixel, and performs decoding. Generate an image. The adding unit 106 stores the generated decoded image in the reference picture memory 109.

The loop filter 107 performs a deblocking filter, a sample adaptive offset (SAO), and an adaptive loop filter (ALF) on the decoded image generated by the adding unit 106.

The prediction parameter memory 108 stores the prediction parameter generated by the encoding parameter determination unit 110 at a predetermined position for each encoding target picture and CU.

The reference picture memory 109 stores the decoded image generated by the loop filter 107 at a predetermined position for each picture to be encoded and each CU.

The encoding parameter determination unit 110 selects one set from among a plurality of sets of encoding parameters. The encoding parameter is a parameter to be encoded that is generated in association with the above-described prediction parameter or the prediction parameter. The predicted image generation unit 101 generates a predicted image P of the PU using each of these encoding parameter sets.

The encoding parameter determination unit 110 calculates a cost value indicating the amount of information and the encoding error for each of a plurality of sets. The cost value is, for example, the sum of a code amount and a square error multiplied by a coefficient λ. The code amount is the information amount of the encoded stream Te obtained by entropy encoding the quantization error and the encoding parameter. The square error is the sum between pixels regarding the square value of the residual value of the residual signal calculated by the subtracting unit 102. The coefficient λ is a real number larger than a preset zero. The encoding parameter determination unit 110 selects a set of encoding parameters that minimizes the calculated cost value. As a result, the entropy encoding unit 104 outputs the selected set of encoding parameters to the outside as the encoded stream Te, and does not output the set of unselected encoding parameters. The encoding parameter determination unit 110 stores the determined encoding parameter in the prediction parameter memory 108.

The prediction parameter encoding unit 111 derives a format for encoding from the parameters input from the encoding parameter determination unit 110 and outputs the format to the entropy encoding unit 104. Deriving the format for encoding is, for example, deriving a difference vector from a motion vector and a prediction vector. Also, the prediction parameter encoding unit 111 derives parameters necessary for generating a prediction image from the parameters input from the encoding parameter determination unit 110 and outputs the parameters to the prediction image generation unit 101. The parameter necessary for generating the predicted image is, for example, a motion vector in units of sub-blocks.

The inter prediction parameter encoding unit 112 derives an inter prediction parameter such as a difference vector based on the prediction parameter input from the encoding parameter determination unit 110. The inter prediction parameter encoding unit 112 derives parameters necessary for generating a prediction image to be output to the prediction image generating unit 101, and an inter prediction parameter decoding unit 303 (see FIG. 5 and the like) derives inter prediction parameters. Some of the configurations are the same as the configuration to be performed. The configuration of the inter prediction parameter encoding unit 112 will be described later.

The intra prediction parameter encoding unit 113 derives a format (for example, MPM_idx, rem_intra_luma_pred_mode) for encoding from the intra prediction mode IntraPredMode input from the encoding parameter determination unit 110.

(Configuration of inter prediction parameter encoding unit)
Next, the configuration of the inter prediction parameter encoding unit 112 will be described. The inter prediction parameter encoding unit 112 is a unit corresponding to the inter prediction parameter decoding unit 303 in FIG. 12, and the configuration is shown in FIG.

The inter prediction parameter encoding unit 112 includes an inter prediction parameter encoding control unit 1121, an AMVP prediction parameter derivation unit 1122, a subtraction unit 1123, a sub-block prediction parameter derivation unit 1125, and a partition mode derivation unit and a merge flag derivation unit (not shown). , An inter prediction identifier deriving unit, a reference picture index deriving unit, a vector difference deriving unit, etc., and a split mode deriving unit, a merge flag deriving unit, an inter prediction identifier deriving unit, a reference picture index deriving unit, and a vector difference deriving unit Respectively derive a PU partition mode part_mode, a merge flag merge_flag, an inter prediction identifier inter_pred_idc, a reference picture index refIdxLX, and a difference vector mvdLX. The inter prediction parameter encoding unit 112 outputs the motion vector (mvLX, subMvLX), the reference picture index refIdxLX, the PU partition mode part_mode, the inter prediction identifier inter_pred_idc, or information indicating these to the prediction image generating unit 101. Also, the inter prediction parameter encoding unit 112 entropy PU partition mode part_mode, merge flag merge_flag, merge index merge_idx, inter prediction identifier inter_pred_idc, reference picture index refIdxLX, prediction vector index mvp_LX_idx, difference vector mvdLX, sub-block prediction mode flag subPbMotionFlag. The data is output to the encoding unit 104.

The inter prediction parameter encoding control unit 1121 includes a merge index deriving unit 11211 and a vector candidate index deriving unit 11212. The merge index derivation unit 11211 compares the motion vector and reference picture index input from the encoding parameter determination unit 110 with the motion vector and reference picture index of the merge candidate PU read from the prediction parameter memory 108, and performs merge An index merge_idx is derived and output to the entropy encoding unit 104. A merge candidate is a reference PU (for example, a reference PU that touches the lower left end, upper left end, and upper right end of the encoding target block) within a predetermined range from the encoding target CU to be encoded. The PU has been processed. The vector candidate index deriving unit 11212 derives a prediction vector index mvp_LX_idx.

When the encoding parameter determination unit 110 determines to use the sub-block prediction mode, the sub-block prediction parameter derivation unit 1125 performs spatial sub-block prediction, temporal sub-block prediction, affine prediction, and matching motion derivation according to the value of subPbMotionFlag. A motion vector and a reference picture index for any sub-block prediction are derived. As described in the description of the image decoding apparatus, the motion vector and the reference picture index are derived by reading out the motion vector and the reference picture index such as the adjacent PU and the reference picture block from the prediction parameter memory 108.

The AMVP prediction parameter derivation unit 1122 has the same configuration as the AMVP prediction parameter derivation unit 3032 (see FIG. 12).

That is, when the prediction mode predMode indicates the inter prediction mode, the motion vector mvLX is input from the encoding parameter determination unit 110 to the AMVP prediction parameter derivation unit 1122. The AMVP prediction parameter derivation unit 1122 derives a prediction vector mvpLX based on the input motion vector mvLX. The AMVP prediction parameter derivation unit 1122 outputs the derived prediction vector mvpLX to the subtraction unit 1123. Note that the reference picture index refIdx and the prediction vector index mvp_LX_idx are output to the entropy encoding unit 104.

The subtraction unit 1123 subtracts the prediction vector mvpLX input from the AMVP prediction parameter derivation unit 1122 from the motion vector mvLX input from the coding parameter determination unit 110 to generate a difference vector mvdLX. The difference vector mvdLX is output to the entropy encoding unit 104.

In addition, a part of the image encoding device 11 and the image decoding device 31 in the above-described embodiment, for example, the entropy decoding unit 301, the prediction parameter decoding unit 302, the loop filter 305, the predicted image generation unit 308, the inverse quantization / inverse transformation. Unit 311, addition unit 312, predicted image generation unit 101, subtraction unit 102, transform / quantization unit 103, entropy encoding unit 104, inverse quantization / inverse transform unit 105, loop filter 107, encoding parameter determination unit 110, The prediction parameter encoding unit 111 may be realized by a computer. In that case, the program for realizing the control function may be recorded on a computer-readable recording medium, and the program recorded on the recording medium may be read by a computer system and executed. Here, the “computer system” is a computer system built in either the image encoding device 11 or the image decoding device 31 and includes hardware such as an OS and peripheral devices. The “computer-readable recording medium” refers to a storage device such as a portable medium such as a flexible disk, a magneto-optical disk, a ROM, a CD-ROM, or a hard disk built in a computer system. Furthermore, the “computer-readable recording medium” is a medium that dynamically holds a program for a short time, such as a communication line when transmitting a program via a network such as the Internet or a communication line such as a telephone line, In such a case, a volatile memory inside a computer system serving as a server or a client may be included and a program that holds a program for a certain period of time. The program may be a program for realizing a part of the functions described above, and may be a program capable of realizing the functions described above in combination with a program already recorded in a computer system.

[Embodiment 2]
(Matching cost in derivation of matching motion)
As described in the first embodiment, for example, there are the following three motion vector derivation methods in the matching motion derivation and BTM.

1. Template matching (TM) for deriving a motion vector by matching a surrounding template of a target block with a motion compensated image derived from a reference image.

2. Bilateral matching (BM) for deriving motion vectors by matching motion compensated images derived from two reference images.

3. Bilateral template matching (BTM) in which a motion vector is updated by matching a template derived from a motion compensated image derived from two reference images and a motion compensated image derived from the reference image.

Also, as described in the first embodiment, in TM, when the matching motion deriving unit 30373 selects an optimal motion vector candidate, matching between a template and a motion compensated image derived from the reference image (Ref0, Ref1) is performed. Cost (costL0, costL1) is calculated. Similarly for the BTM, the BTM processing unit 3038 derives a matching cost (costL0, costL1) between the template and the motion compensated image derived from the reference images (Ref0, Ref1). The matching cost is a cost corresponding to the goodness of matching (error amount, difference). In BM, the matching cost for each of the motion compensated images derived from the two reference images is not derived.

For example, in TM, the matching motion deriving unit 30373 derives costL0 and costL1 by the following expressions.

costL0 = Σ | motion compensation image [x] [y] | on template image [x] [y] -Ref0
costL1 = Σ | motion compensation image [x] [y] | on template image [x] [y] -Ref1
Note that x and y are coordinates in the template, and when the width and height of the template are blkW and blkH, x = 0..blkW −1 and y = 0..blkH −1 are taken. Σ is an operation that takes the sum of x and y in the above range. When multiple regions are available as templates (for example, the left region and the upper region), the sum of the costs in each template may be used as the matching cost (costL0 of the reference image Ref0 of the motion compensated image predSamplesL0 [x] [y]) , CostL1 of the reference image Ref1 of the motion compensated image predSamplesL1 [x] [y]). Note that the template image and the motion compensated images Ref0 and Ref1 may be stored once in a one-dimensional temporal buffer and then compared. In this case, the above summation is performed not on the subscripts x and y but on a one-dimensional buffer position (for example, i).

For example, costL0 and costL1 are derived in BTM. Since the BTM template image is predBTM [] [] described above, the BTM processing unit 3038 derives costL0 and costL1 according to the following equations.

costL0 = Σ | motion compensation image [x] [y] | on template image predBTM [x] [y] -Ref0
costL1 = Σ | motion compensation image [x] [y] | on template image predBTM [x] [y] -Ref1
X and y are the coordinates in the block, and when the block width and height are blkW and blkH, x = 0..blkW −1 and y = 0..blkH −1 are taken. Σ is an operation that takes the sum of x and y in the above range.

In the above, the sum of absolute differences SAD (Sum of Absolute Difference) between the pixel value of the template and the pixel value of the reference image is used to calculate the matching cost costLX for each reference image RefX (X = 0 or 1) in TM and BTM As an example, square error sum SSD (Sum of Square Differenec) or SATD (Sum of Absolute Transformed Differences) may be used in addition to SAD. The derivation formula of SAD corresponding to the cost costLX for the reference image RefX is as described above.
costLX = SAD = Σ | Motion compensated image [x] [y] | on template image [x] [y] -RefX
It is.

Also, as shown in the following equation, the matching cost costLX for each reference image RefX (X = 0 or 1) in TM and BTM is set to the above SAD as an initial vector (IniMVx, IniMVy) and a motion vector candidate (MVx, MVy) ) May be derived by adding the difference to.

(Outline of this embodiment)
For example, in the configuration described in the first embodiment, in TM and BTM, the weight prediction unit 3094 converts the weight prediction image predSamples [x] [y], the motion compensation image predSamplesL0 [x] [y], and the motion compensation image predSamplesL1 [ Derived from the average value of x] [y].

That is, in the configuration illustrated in the first embodiment, the weight prediction unit 3094 derives a weight prediction image without considering the matching cost of each motion compensation image.

The weight prediction unit 3094 according to the present embodiment derives weight prediction images predSamples [x] [y] in TM and BTM in consideration of the matching cost of each motion compensation image.

According to such a configuration, the weight prediction unit 3094 can derive weight prediction images predSamples [x] [y] with high accuracy reflecting the matching cost of each motion compensation image.

Specifically, the BTM processing unit (matching processing unit) 3038 and the matching motion derivation unit (matching processing unit) 30373 perform matching processing using bilateral template matching (BTM) or template matching (TM), and A difference (matching cost) between a predicted image candidate (motion compensation image and interpolation image) derived from a reference picture and a template is derived. The weight prediction unit 3094 derives a predicted image from a plurality of predicted image candidates predicted for each reference picture. The weight prediction unit 3094 sets a weighting coefficient (weight) for each prediction image candidate according to the difference between the template and the prediction image candidate, and derives the prediction image using the weighted prediction image candidate. To do.

According to the above configuration, the matching cost of the predicted image candidate derived in bilateral template matching or template matching can be effectively used for deriving the predicted image.

(Derivation of weight prediction image using matching cost in BTM)
(Inter prediction image generation unit 309)
Next, the inter prediction image generation unit 309 according to the present embodiment will be described with reference to FIG. FIG. 22 is a block diagram illustrating an example of a main configuration of the inter predicted image generation unit 309 according to the present embodiment. As illustrated in FIG. 22, the inter predicted image generation unit 309 includes a motion compensation unit 3091 and a weight prediction unit 3094. Since the motion compensation unit 3091 has the same configuration as that described in the first embodiment, detailed description thereof is omitted here.

(Weight predictor 3094)
The weight prediction unit 3094 receives the matching cost (costL0, costL1) derived in the BTM from the BTM processing unit 3038. The weight prediction unit 3094 weights the motion compensated images (predSamplesL0 [x] [y], predSamplesL1 [x] [y]) according to the matching costs (costL0, costL1). The weight prediction unit 3094 derives the weight prediction image predSamples [x] [y] using the weighted motion compensation images (predSamplesL0 [x] [y], predSamplesL1 [x] [y]).

(Weight derivation example and prediction image generation)
Here, an example of the weight (weight) derived by the weight prediction unit 3094 will be described.

The weight prediction unit 3094 compares costL0 and costL1.

When costL0 is smaller than a predetermined level compared to costL1 (for example, costL0 * 2 <costL1), the weight prediction unit 3094 uses the weight w0 of predSamplesL0 [x] [y] from the weight w1 of predSamplesL1 [x] [y] Also make it bigger.

Further, when costL1 is smaller than a predetermined level compared with costL0 (for example, costL1 * 2 <costL0), the weight prediction unit 3094 uses the weight w1 of predSamplesL1 [x] [y] as the weight of predSamplesL0 [x] [y]. Make it larger than w0.

In other words, the weight prediction unit 3094 sets a larger value of the weighting coefficient to the motion compensation image having a smaller difference from the template than the motion compensation image having a larger difference (matching cost) from the template.

According to the above configuration, a large weight is attached to a motion compensated image with a small amount of error from the template. Therefore, the predicted image generation device can derive a predicted image with high accuracy.

In other cases, the weight prediction unit 3094 sets the weight w1 of predSamplesL1 [x] [y] and the weight w0 of predSamplesL0 [x] [y] to the same size.

The above derivation example of the weight can be expressed by the following equation.

w0 = w1 = Winit
if (costL0 * 2 <costL1) w0 = Wh; w1 = 16-w0
else if (costL1 * 2 <costL0) w0 = Wl; w1 = 16-w0
For example, Winit = 4, Wh = 12, and Wl = 4 can be set.

Next, as shown in the following equation, the weight prediction unit 3094 weights the motion compensated image using the derived weight and derives the weight predicted image predSamples [X] [Y].

predSamples [x] [y] = (w0 * predSamplesL0 [x] [y] + w1 + predSamplesL1 [x] [y] + roundoffset) >> shift
roundoffset = 1 << (shift-1)
shift = log2 (w0 + w1)
Note that an offset addition process or a clip process may be further added as described below.

predSamples [x] [y] = Clip3 (0, (1 << bitDepth)-1, (predSamplesL0 [x] [y] * w0 + predSamplesL1 [x] [y] * w1 + ((o0 + o1 + 1) << log2WD)) >> (log2WD + 1))
(Other examples of weights)
Another example of the weight (weight) derived by the weight prediction unit 3094 will be described.

The weight prediction unit 3094 sets a weighting coefficient for the motion compensation image according to the ratio of the matching cost of the motion compensation image.

According to the above configuration, a large value is weighted to a motion compensated image with a small amount of error from the template. Therefore, the inter prediction image generation unit 309 (prediction image generation device) can derive a prediction image with high accuracy.

In this example, the weight prediction unit 3094 derives a weight from the cost ratio between costL0 and costL1. The details of the weight derivation example using the cost ratio are shown below.

(Example 1)
The weight prediction unit 3094 derives a weight according to the cost ratio as shown in the following equation. For example, the weight prediction unit 3094 derives the weight so that the weight of the reference image with the smaller cost out of the costL0 of the reference image Ref0 and the costL1 of the reference image Ref1 is larger than the weight of the reference image with the larger cost. To do. That is, when costL0 is smaller than costL1, weights w0 and w1 are derived so that weight w0 ≧ weight w1. For example, the ratio of costL0 to total costL0 and costL1 is used.

w1 = (1 << shift) * (costL0 / (costL0 + costL1))
w0 = (1 << shift)-w1
In addition, you may derive | lead-out in order of w0, w1 instead of deriving in order of w1, w0 (the same result is obtained).

w0 = (1 << shift) * (costL1 / (costL0 + costL1))
w1 = (1 << shift)-w0
(Example 2)
Similar to Example 1, the weight prediction unit 3094 derives a weight according to the cost ratio as shown in the following equation. However, the division of “(costL0 / (costL0 + costL1)” in Example 1 is a table division.

w1 = (1 << shift) * (costL0 * InvTbl [(costL0 + costL1)] >> invTblShift)
w0 = (1 << shift)-w1
Reciprocal table InvTbl [x] = (1 << invTblShift + x / 2) / x
Where x = 0 .. (1 << invTblShift) -1.

As the reciprocal table, for example, it is appropriate to have a 32768 element table with invTblShift = 15 and x = 0..32767.

Of course, the same result can be obtained by the following equation.

w0 = (1 << shift) * (costL1 * InvTbl [(costL0 + costL1)] >> invTblShift)
w1 = (1 << shift)-w0
In the following Examples 3 to 5, description of such obvious modifications will be omitted.

(Example 3)
As shown in the following equation, the weight prediction unit 3094 adds a constant C to the numerator and denominator in “costL0 * InvTbl [(costL0 + costL1)]” in Example 2 rather than weight derivation according to the cost ratio. Determine the weight ratio at a moderate rate.

w1 = (1 << shift) * ((costL0 + C) * InvTbl [(costL0 + costL1 + C)] >> invTblShift)
w0 = (1 << shift)-w1
Reciprocal table IInvTbl [x] = (1 << invTblShift + x / 2) / x
(Example 4)
As shown in the following equation, the weight prediction unit 3094 derives the ratio K between the numerator and the denominator of “costL0 * InvTbl [(costL0 + costL1)]” in Example 3, and then performs weighting by table subtraction (mapping) from K. Determine the ratio.

K = (1 << shiftK) * ((costL0 + C) * InvTbl [(costL0 + costL1 + C)] >> invTblShift)
w1 = weightTbl [K]
w0 = (1 << shift)-w1
Reciprocal table IInvTbl [x] = (1 << invTblShift + x / 2) / x
For example, the weight mapping table weightTbl [] = {2, 2, 3, 3, 3, 3, 4, 4, 4, 4, 4, 5, 5, 5, 5, 6}
shiftK = 4 (in this case, the range of K is K = 0 .. 1 << shiftK-1, that is, K = 0..7)
shift = 3
It can be.

(Example 5)
Another example of derivation of weights will be described with reference to FIG. FIG. 23 is a pseudo code showing an example of a weight derivation process by the weight prediction unit 3094.

The weight prediction unit 3094 sets the weighting coefficient set to the motion compensation image to a constant value when the sum of the matching costs of the motion compensation image is smaller than a predetermined value.

According to the above configuration, a large value is weighted to the motion compensated image with a small amount of error from the template. Therefore, the inter prediction image generation unit 309 can derive a prediction image with high accuracy.

Specifically, as shown in the equation of FIG. 23, when the sum of costL0 and costL1 is smaller than the threshold TH_K, the weight prediction unit 3094 may always set the weight ratio between w0 and w1 to 1: 1.

Also, if costL0 and costL1 satisfy costL0 * 2 <costL1, w0 = 12; w1 = 16-w0 may be set.

Also, if costL0 and costL1 satisfy costL1 * 2 <costL0, w0 = 4; w1 = 16-w0 may be set.

In the equation shown in FIG. 23, width and height indicate the width and height of the target block (= predicted block), and TH_K indicates a constant serving as a threshold value of the sum of costL0 and costL1. For example, TH_K = 2 may be set.

(Derivation of weight prediction image using matching cost in TM)
(Inter prediction image generation unit 309)
Next, the inter prediction image generation unit 309 according to the present embodiment will be described with reference to FIG. FIG. 24 is a block diagram illustrating an example of a main configuration of the inter predicted image generation unit 309 according to the present embodiment. As illustrated in FIG. 24, the inter predicted image generation unit 309 includes a motion compensation unit 3091 and a weight prediction unit 3094. Since the motion compensation unit 3091 has the same configuration as that described in the first embodiment, detailed description thereof is omitted here.

(Weight predictor 3094)
The weight prediction unit 3094 receives the matching cost (costL0, costL1) derived in the TM from the matching motion derivation unit 30373. The weight prediction unit 3094 weights the motion compensated images (predSamplesL0 [x] [y], predSamplesL1 [x] [y]) according to the matching costs (costL0, costL1). The weight prediction unit 3094 derives the weight prediction image predSamples [x] [y] using the weighted motion compensation images (predSamplesL0 [x] [y], predSamplesL1 [x] [y]) as shown in the following equation. .

predSamples [x] [y] = (w0 * predSamplesL0 [x] [y] + w1 + predSamplesL1 [x] [y] + roundoffset) >> shift
Note that the derivation example described in the above “derivation of weight prediction image using matching cost in BTM” can be applied to the weight derivation performed by the weight prediction unit 3094. Omitted.

[Embodiment 3]
(BIO ON / OFF control using cost)
The motion compensation unit 3091 according to the present embodiment performs motion compensation using Bi-directional optical flow (bi-predictive gradient change: BIO) prediction that performs motion correction by applying a gradient correction term to bi-prediction (BiPred). It has a mode to do.

The motion compensation unit 3091 performs BIO prediction ON / OFF control according to the matching cost in BTM. For example, the motion compensation unit 3091 receives the matching cost (costL0, costL1) from the BTM processing unit 3038. The motion compensation unit 3091 controls ON / OF of BIO prediction according to the cost ratio between costL0 and costL1.

In other words, the BTM processing unit 3038 performs a matching process using bilateral template matching (BTM), and derives a difference (matching cost) between the motion compensated image predicted from the template in a plurality of reference pictures and the template. .

The motion compensation unit 3091 performs motion compensation by bi-predictive gradient change (BIO) prediction that performs motion compensation by applying a gradient correction term to bilateral template matching (BTM). The motion compensation unit 3091 controls ON / OFF of BIO prediction according to a ratio of a difference (matching cost) between a motion compensated image predicted from a template in a plurality of reference pictures and the template.

According to the above configuration, the matching cost of the motion compensated image derived in the bilateral template matching can be effectively used for deriving the predicted image. Then, ON / OFF of the BIO prediction can be controlled according to the matching cost ratio of the motion compensation image.

Referring to FIG. 25, the control of BIO prediction ON / OFF will be described in detail. FIG. 25 is a diagram illustrating an example of pseudo code indicating processing of the motion compensation unit 3091. As shown in FIG. 25, the motion compensation unit 3091 turns off BIO prediction when the cost ratio is large, and turns on BIO prediction in other cases. The determination of the cost ratio may be performed, for example, by comparing costL0 and costL1 as shown in the following equation.

Judgment value with large cost ratio = costL0> costL1 * BIO_TH || costL1> costL0 * BIO_TH
BIO_TH may be 4, for example.

In other words, the motion compensation unit 3091 controls the BIO prediction to be OFF when the matching cost ratio is larger than a predetermined value.

For example, if the ratio of matching costs is larger than a predetermined value, it is contrary to the BIO hypothesis that the object does not deform in the local region and only performs translation (matching cost when only translation is performed). Does not change significantly depending on the reference image). Therefore, BIO prediction is inappropriate.

According to the above configuration, when the matching cost ratio is larger than a predetermined value, the BIO prediction can be controlled to be OFF.

As shown in FIG. 25, when the cost (costL0, costL1) is small, the motion compensation unit 3091 may turn on the BIO even if the cost ratio is large.

For example, the following formula may be used for determining that the cost is large, using the block width and height as blkW and blkH.

Cost is high = costL0 + costL1> blkW * blkH * TH_K (for example, TH_K = 8)
In other words, the motion compensation unit 3091 controls BIO prediction to be ON when the sum of matching costs is smaller than a predetermined value.

For example, when the sum of the matching costs is smaller than a predetermined value, it matches the BIO hypothesis that the object does not deform in the local region and only performs translation (matching is performed when only translation is performed). The cost is small). Therefore, BIO prediction is preferred.

According to the above configuration, when the sum of the matching costs is smaller than a predetermined value, the BIO prediction can be controlled to be ON.

[Additional Notes]
Moreover, you may implement | achieve part or all of the image coding apparatus 11 in the embodiment mentioned above, and the image decoding apparatus 31 as integrated circuits, such as LSI (Large Scale Integration). Each functional block of the image encoding device 11 and the image decoding device 31 may be individually made into a processor, or a part or all of them may be integrated into a processor. Further, the method of circuit integration is not limited to LSI, and may be realized by a dedicated circuit or a general-purpose processor. Further, in the case where an integrated circuit technology that replaces LSI appears due to progress in semiconductor technology, an integrated circuit based on the technology may be used.

As described above, the embodiment of the present invention has been described in detail with reference to the drawings. However, the specific configuration is not limited to the above, and various design changes and the like can be made without departing from the scope of the present invention. It is possible to

[Application example]
The image encoding device 11 and the image decoding device 31 described above can be used by being mounted on various devices that perform transmission, reception, recording, and reproduction of moving images. The moving image may be a natural moving image captured by a camera or the like, or an artificial moving image (including CG and GUI) generated by a computer or the like.

First, it will be described with reference to FIG. 26 that the image encoding device 11 and the image decoding device 31 described above can be used for transmission and reception of moving images.

(A) of FIG. 26 is a block diagram showing a configuration of a transmission device PROD_A in which the image encoding device 11 is mounted. As illustrated in (a) of FIG. 26, the transmission apparatus PROD_A modulates a carrier wave with an encoding unit PROD_A1 that obtains encoded data by encoding a moving image, and with the encoded data obtained by the encoding unit PROD_A1. Thus, a modulation unit PROD_A2 that obtains a modulation signal and a transmission unit PROD_A3 that transmits the modulation signal obtained by the modulation unit PROD_A2 are provided. The above-described image encoding device 11 is used as the encoding unit PROD_A1.

Transmission device PROD_A, as a source of moving images to be input to the encoding unit PROD_A1, a camera PROD_A4 that captures moving images, a recording medium PROD_A5 that records moving images, an input terminal PROD_A6 for inputting moving images from the outside, and An image processing unit A7 that generates or processes an image may be further provided. FIG. 26A illustrates a configuration in which the transmission apparatus PROD_A includes all of these, but some of them may be omitted.

Note that the recording medium PROD_A5 may be a recording of a non-encoded moving image, or a recording of a moving image encoded by a recording encoding scheme different from the transmission encoding scheme. It may be a thing. In the latter case, a decoding unit (not shown) for decoding the encoded data read from the recording medium PROD_A5 in accordance with the recording encoding method may be interposed between the recording medium PROD_A5 and the encoding unit PROD_A1.

(B) of FIG. 26 is a block diagram illustrating a configuration of the receiving device PROD_B in which the image decoding device 31 is mounted. As shown in FIG. 26B, the receiving device PROD_B includes a receiving unit PROD_B1 that receives a modulated signal, a demodulating unit PROD_B2 that obtains encoded data by demodulating the modulated signal received by the receiving unit PROD_B1, and a demodulator. A decoding unit PROD_B3 that obtains a moving image by decoding the encoded data obtained by the unit PROD_B2. The above-described image decoding device 31 is used as the decoding unit PROD_B3.

The receiving device PROD_B is a display destination PROD_B4 for displaying a moving image, a recording medium PROD_B5 for recording a moving image, and an output terminal for outputting the moving image to the outside as a supply destination of the moving image output by the decoding unit PROD_B3 PROD_B6 may be further provided. FIG. 26B illustrates a configuration in which the reception apparatus PROD_B includes all of these, but a part of the configuration may be omitted.

Note that the recording medium PROD_B5 may be used for recording a non-encoded moving image, or is encoded using a recording encoding method different from the transmission encoding method. May be. In the latter case, an encoding unit (not shown) for encoding the moving image acquired from the decoding unit PROD_B3 according to the recording encoding method may be interposed between the decoding unit PROD_B3 and the recording medium PROD_B5.

Note that the transmission medium for transmitting the modulation signal may be wireless or wired. Further, the transmission mode for transmitting the modulated signal may be broadcasting (here, a transmission mode in which the transmission destination is not specified in advance) or communication (here, transmission in which the transmission destination is specified in advance). Refers to the embodiment). That is, the transmission of the modulation signal may be realized by any of wireless broadcasting, wired broadcasting, wireless communication, and wired communication.

For example, a terrestrial digital broadcast broadcasting station (broadcasting equipment, etc.) / Receiving station (such as a television receiver) is an example of a transmitting device PROD_A / receiving device PROD_B that transmits and receives a modulated signal by wireless broadcasting. A broadcasting station (such as broadcasting equipment) / receiving station (such as a television receiver) of cable television broadcasting is an example of a transmitting device PROD_A / receiving device PROD_B that transmits and receives a modulated signal by cable broadcasting.

In addition, a server (workstation, etc.) / Client (television receiver, personal computer, smartphone, etc.) such as a VOD (Video On Demand) service or a video sharing service using the Internet is a transmission device that transmits and receives modulated signals via communication. This is an example of PROD_A / receiving device PROD_B (normally, either a wireless or wired transmission medium is used in a LAN, and a wired transmission medium is used in a WAN). Here, the personal computer includes a desktop PC, a laptop PC, and a tablet PC. The smartphone also includes a multi-function mobile phone terminal.

In addition to the function of decoding the encoded data downloaded from the server and displaying it on the display, the video sharing service client has a function of encoding a moving image captured by the camera and uploading it to the server. That is, the client of the video sharing service functions as both the transmission device PROD_A and the reception device PROD_B.

Next, the fact that the above-described image encoding device 11 and image decoding device 31 can be used for recording and reproduction of moving images will be described with reference to FIG.

FIG. 27A is a block diagram showing a configuration of a recording apparatus PROD_C in which the above-described image encoding device 11 is mounted. As shown in (a) of FIG. 27, the recording apparatus PROD_C includes an encoding unit PROD_C1 that obtains encoded data by encoding a moving image, and the encoded data obtained by the encoding unit PROD_C1 on a recording medium PROD_M. A writing unit PROD_C2 for writing. The above-described image encoding device 11 is used as the encoding unit PROD_C1.

The recording medium PROD_M may be of a type built into the recording device PROD_C, such as (1) HDD (Hard Disk Drive) or SSD (Solid State Drive), or (2) SD memory. It may be of the type connected to the recording device PROD_C, such as a card or USB (Universal Serial Bus) flash memory, or (3) DVD (Digital Versatile Disc) or BD (Blu-ray Disc: registration) Or a drive device (not shown) built in the recording device PROD_C.

In addition, the recording device PROD_C is a camera PROD_C3 that captures moving images as a source of moving images to be input to the encoding unit PROD_C1, an input terminal PROD_C4 for inputting moving images from the outside, and a reception for receiving moving images A unit PROD_C5 and an image processing unit PROD_C6 for generating or processing an image may be further provided. FIG. 27A illustrates a configuration in which the recording apparatus PROD_C includes all of these, but some of them may be omitted.

The receiving unit PROD_C5 may receive a non-encoded moving image, or may receive encoded data encoded by a transmission encoding scheme different from the recording encoding scheme. You may do. In the latter case, a transmission decoding unit (not shown) that decodes encoded data encoded by the transmission encoding method may be interposed between the reception unit PROD_C5 and the encoding unit PROD_C1.

Examples of such a recording device PROD_C include a DVD recorder, a BD recorder, an HDD (Hard Disk Drive) recorder, and the like (in this case, the input terminal PROD_C4 or the receiver PROD_C5 is a main source of moving images). . In addition, a camcorder (in this case, the camera PROD_C3 is a main source of moving images), a personal computer (in this case, the receiving unit PROD_C5 or the image processing unit C6 is a main source of moving images), a smartphone (this In this case, the camera PROD_C3 or the reception unit PROD_C5 is a main source of moving images), and the like is also an example of such a recording apparatus PROD_C.

(B) of FIG. 27 is a block diagram showing a configuration of a playback device PROD_D equipped with the image decoding device 31 described above. As shown in (b) of FIG. 27, the playback device PROD_D reads a moving image by decoding a read unit PROD_D1 that reads encoded data written on the recording medium PROD_M and a read unit PROD_D1 that reads the encoded data. And a decoding unit PROD_D2 to obtain. The above-described image decoding device 31 is used as the decoding unit PROD_D2.

The recording medium PROD_M may be of the type built into the playback device PROD_D, such as (1) HDD or SSD, or (2) such as an SD memory card or USB flash memory. It may be of the type connected to the playback device PROD_D, or (3) may be loaded into a drive device (not shown) built in the playback device PROD_D, such as a DVD or BD. Good.

In addition, the playback device PROD_D has a display unit PROD_D3 that displays a moving image as a supply destination of the moving image output by the decoding unit PROD_D2, an output terminal PROD_D4 that outputs the moving image to the outside, and a transmission unit that transmits the moving image. PROD_D5 may be further provided. FIG. 27B illustrates a configuration in which the playback apparatus PROD_D includes all of these, but some of them may be omitted.

The transmission unit PROD_D5 may transmit a non-encoded moving image, or transmits encoded data encoded by a transmission encoding scheme different from the recording encoding scheme. You may do. In the latter case, it is preferable to interpose an encoding unit (not shown) that encodes a moving image using a transmission encoding method between the decoding unit PROD_D2 and the transmission unit PROD_D5.

Examples of such a playback device PROD_D include a DVD player, a BD player, and an HDD player (in this case, an output terminal PROD_D4 to which a television receiver or the like is connected is a main moving image supply destination). . In addition, a television receiver (in this case, the display PROD_D3 is a main supply destination of moving images), a digital signage (also referred to as an electronic signboard or an electronic bulletin board), and the display PROD_D3 or the transmission unit PROD_D5 is the main supply of moving images Desktop PC (in this case, output terminal PROD_D4 or transmission unit PROD_D5 is the main video source), laptop or tablet PC (in this case, display PROD_D3 or transmission unit PROD_D5 is video) A smartphone (which is a main image supply destination), a smartphone (in this case, the display PROD_D3 or the transmission unit PROD_D5 is a main moving image supply destination), and the like are also examples of such a playback device PROD_D.

(Hardware implementation and software implementation)
Each block of the image decoding device 31 and the image encoding device 11 described above may be realized in hardware by a logic circuit formed on an integrated circuit (IC chip), or may be a CPU (Central Processing Unit). You may implement | achieve by software using.

In the latter case, each of the above devices includes a CPU that executes instructions of a program that realizes each function, a ROM (Read Memory) that stores the program, a RAM (RandomAccess Memory) that develops the program, the program, and various data. A storage device (recording medium) such as a memory for storing the. The object of the embodiment of the present invention is a record in which the program code (execution format program, intermediate code program, source program) of the control program for each of the above devices, which is software that realizes the above-described functions, is recorded in a computer-readable manner This can also be achieved by supplying a medium to each of the above devices, and reading and executing the program code recorded on the recording medium by the computer (or CPU or MPU).

Examples of the recording medium include tapes such as magnetic tapes and cassette tapes, magnetic disks such as floppy (registered trademark) disks / hard disks, CD-ROMs (Compact Disc Read-Only Memory) / MO discs (Magneto-Optical discs). ) / MD (Mini Disc) / DVD (Digital Versatile Disc) / CD-R (CD Recordable) / Blu-ray Disc (Blu-ray Disc: registered trademark) and other optical disks, IC cards (including memory cards) / Cards such as optical cards, Mask ROM / EPROM (Erasable Programmable Read-Only Memory) / EEPROM (Electrically Erasable and Programmable Read-Only Memory: registered trademark) / Semiconductor memories such as flash ROM, or PLD (Programmable logic device ) Or FPGA (Field Programmable Gate Gate Array) or the like.

Further, each of the above devices may be configured to be connectable to a communication network, and the program code may be supplied via the communication network. The communication network is not particularly limited as long as it can transmit the program code. For example, Internet, intranet, extranet, LAN (Local Area Network), ISDN (Integrated Services Digital Network), VAN (Value-Added Network), CATV (Community Area Antenna / television / Cable Television), Virtual Private Network (Virtual Private Network) Network), telephone line network, mobile communication network, satellite communication network, and the like. The transmission medium constituting the communication network may be any medium that can transmit the program code, and is not limited to a specific configuration or type. For example, IEEE (Institute of Electrical and Electronic Engineers) 1394, USB, power line carrier, cable TV line, telephone line, ADSL (Asymmetric Digital Subscriber Line) line, etc. wired such as IrDA (Infrared Data Association) or remote control , BlueTooth (registered trademark), IEEE802.11 wireless, HDR (High Data Rate), NFC (Near Field Communication), DLNA (Digital Living Network Alliance: registered trademark), mobile phone network, satellite line, terrestrial digital broadcasting network, etc. It can also be used wirelessly. The embodiment of the present invention can also be realized in the form of a computer data signal embedded in a carrier wave in which the program code is embodied by electronic transmission.

The embodiments of the present invention are not limited to the above-described embodiments, and various modifications are possible within the scope of the claims. That is, embodiments obtained by combining technical means appropriately modified within the scope of the claims are also included in the technical scope of the present invention.

(Cross-reference of related applications)
This application claims the benefit of priority over the Japanese patent application filed on March 21, 2017: Japanese Patent Application No. 2017-055101. Included in this document.

303 Inter prediction parameter decoding unit (predicted image generation device)
30373 Matching motion deriving unit (matching processing unit)
3038 BTM processing unit (matching processing unit)
309 inter prediction image generation unit (prediction image generation device)
3091 Motion compensation unit 3094 Weight prediction unit (prediction unit)
31 Image decoding apparatus 11 Image encoding apparatus

Claims

A prediction image generation device that generates a prediction image used for encoding or decoding a moving image,
A matching processing unit that performs bilateral template matching (BTM) or template matching (TM) to perform a matching process, and derives a cost of the predicted image candidate derived from the template in a plurality of reference pictures and the template,
A prediction unit for deriving a predicted image from a plurality of predicted image candidates predicted for each reference picture,
The prediction unit sets a weighting coefficient for each prediction image candidate according to the cost of the template and the prediction image candidate, and derives the prediction image using the weighted prediction image candidate. A predicted image generation apparatus.
The said prediction part sets the said weighting coefficient of a larger value to the said prediction image candidate with the said cost with the said template smaller than the said prediction image candidate with the said large cost with the said template, It is characterized by the above-mentioned. The predicted image generation device according to 1.
2. The predicted image generation apparatus according to claim 1, wherein the prediction unit sets a weighting coefficient for the predicted image candidate in accordance with the cost ratio of the predicted image candidate.
4. The prediction unit according to claim 1, wherein when the sum of the costs of the prediction image candidates is smaller than a predetermined value, the prediction unit sets a weighting coefficient to be set to the prediction image candidate as a constant value. 5. The predicted image generation device according to claim 1.
A prediction image generation device that generates a prediction image used for encoding or decoding a moving image,
A matching processing unit that performs a matching process using bilateral template matching (BTM) and derives a cost of the predicted image candidate predicted from the template in a plurality of reference pictures and the template;
A motion compensation unit that performs motion compensation by bi-predictive gradient change (BIO) prediction that performs motion compensation by applying a gradient correction term to bilateral template matching,
The motion compensation unit controls ON / OFF of BIO prediction according to a cost ratio between the predicted image candidate predicted from a template in a plurality of reference pictures and the template. .
The predicted image generation apparatus according to claim 5, wherein the motion compensation unit controls the BIO prediction to be OFF when the cost ratio is larger than a predetermined value.
The predicted image generation apparatus according to claim 5 or 6, wherein the motion compensation unit controls BIO prediction to be ON when the sum of the costs is smaller than a predetermined value.
8. The predicted image generation apparatus according to claim 1, wherein the cost is a sum of absolute differences between a template and a predicted image candidate.
The said cost is represented by the function of the absolute difference sum of a template and a prediction image candidate, and the difference of the initial vector of a template and a prediction image candidate, and a final motion vector, It is characterized by the above-mentioned. Prediction image generation apparatus.
A moving picture decoding apparatus comprising the predicted image generation apparatus according to any one of claims 1 to 9.
A moving picture encoding apparatus comprising the predicted image generation apparatus according to any one of claims 1 to 9.