Method, apparatus and device for determining partitioning type in video encoding, and storage medium

Abstract

Embodiments of the present application provide a method, apparatus and device for determining a partitioning type in video encoding, and a storage medium. The method comprises: upon completion of a binary tree partitioning decision for a current encoding region, acquiring partitioning type information corresponding to a first partitioning depth, the number of encoding units corresponding to a second partitioning depth, and a second optimal partitioning type for each encoding unit; then, comparing the number of encoding units with a preset quantity threshold; and finally, when a first optimal partitioning type is binary tree partitioning in a preset partitioning direction, the number of encoding units is greater than or equal to the preset quantity threshold, and the second optimal partitioning type for each encoding unit is a partitioning type in the preset partitioning direction, skipping tree partitioning decisions for the current encoding region that comprise other partitioning directions. The present solution achieves skipping of partitioning type decisions to be executed having low probabilities of being optimal, while minimizing loss in encoding quality, thereby improving video encoding efficiency.

Description

Method, device and equipment for partition type decision of video coding and storage medium

[0001] The present application claims priority to the Chinese patent application No. 202410848399.2, filed on June 27, 2024, the content of which is incorporated herein by reference in its entirety. TECHNICAL FIELD

[0002] Embodiments of the present application relate to the field of video coding technology, in particular to a method, device and equipment for partition type decision of video coding and storage medium. BACKGROUND

[0003] In a new generation of encoder standard, multiple recursive partition structures can be supported, including no partition, quad-tree partition, binary-tree partition and ternary-tree partition, wherein the binary-tree partition includes vertical binary-tree and horizontal binary-tree, and the ternary-tree partition includes vertical ternary-tree and horizontal ternary-tree. In the partition process, the coding unit can select any partition tree, but once the coding unit selects a multi-type tree, the subsequent nodes can only select a multi-type tree, and there is a limit to the maximum multi-type tree depth. The new generation of encoder standard can decide, within a limited depth range, according to the rate-distortion cost based on the video content adaptation, which layer the current region needs to be partitioned into and which partition is used. Generally, an encoding tree unit is taken as a root node, a no partition type decision is first made, and then it is determined whether to preferentially perform quad-tree partition or binary-tree partition according to the value of the partition flag. In the case of preferentially performing binary-tree partition, horizontal binary-tree partition is first performed, and then multi-type tree partition is recursively performed on the partition result according to the maximum partition allowed depth; after the partition ends, vertical binary-tree partition is continued at the previous layer, and multi-type tree partition is recursively performed on the partition result according to the rules; then ternary-tree partition is continued at the previous layer, and after the binary-tree partition and ternary-tree partition are completed; then quad-tree partition is performed, and multi-type tree partition is recursively performed on the partition result according to the rules; finally, the partition structure with the minimum rate-distortion cost is taken as the final partition result of the encoding tree unit.

[0004] Although the introduction of the new partition structure in the new generation of encoder standard brings better coding benefits, more partition type decisions need to be performed, which increases the computational complexity of the video encoder, reduces the coding speed, affects the video coding efficiency, and needs to be improved. SUMMARY

[0005] The embodiment of the present application provides a video coding partition type decision method, device, equipment and storage medium, solves the problems of increasing the calculation complexity of a video encoder, reducing the coding speed and influencing the video coding efficiency due to the need of performing more partition type decisions, realizes the reasonable determination of whether to perform skipping on the tree type partition decision of the current coding area containing other partition directions by referring to the optimal coding unit partition information after binary tree partition decision, skips the to-be-partitioned type decision with a smaller optimal probability on the premise of less loss of coding quality, reduces the calculation complexity of the video encoder, improves the coding speed, and thus improves the video coding efficiency.

[0006] In a first aspect, the embodiment of the present application provides a video coding partition type decision method, which comprises the following steps:

[0007] In the case that the binary tree partition decision of the current coding area is completed, the partition type information corresponding to a first partition depth, the number of coding units corresponding to a second partition depth and the second optimal partition type of each coding unit are acquired, wherein the partition type information comprises a first optimal partition type, and the second partition depth is greater than the first partition depth;

[0008] The number of coding units is compared with a preset number threshold;

[0009] In the case that the first optimal partition type is a binary tree partition of a preset partition direction, the number of coding units is greater than or equal to the preset number threshold, and the second optimal partition type of each coding unit is a partition type of a preset partition direction, the tree type partition decision of the current coding area containing other partition directions is skipped.

[0010] In a second aspect, the embodiment of the present application further provides a video coding partition type decision device, which comprises:

[0011] The partition information acquisition module is configured to acquire, in the case that the binary tree partition decision of the current coding area is completed, the partition type information corresponding to a first partition depth, the number of coding units corresponding to a second partition depth and the second optimal partition type of each coding unit, wherein the partition type information comprises a first optimal partition type, and the second partition depth is greater than the first partition depth;

[0012] The unit number comparison module is configured to compare the number of coding units with a preset number threshold;

[0013] The partition decision skipping module is configured to skip the tree partition decision of other partition directions of the current coding region when the first optimal partition type is the binary tree partition of the preset partition direction, the number of the coding units is greater than or equal to the preset number threshold, and the second optimal partition type of each coding unit is the preset partition direction.

[0014] In a third aspect, an apparatus for partition type decision of video coding is provided. The apparatus includes:

[0015] one or more processors;

[0016] a storage configured to store one or more programs,

[0017] When the one or more programs are executed by the one or more processors, the one or more processors implement the method for partition type decision of video coding provided by the embodiments of the present disclosure.

[0018] In a fourth aspect, a non-transitory storage medium storing computer executable instructions is provided. When the computer executable instructions are executed by a computer processor, the computer processor is configured to perform the method for partition type decision of video coding provided by the embodiments of the present disclosure.

[0019] In a fifth aspect, a computer program product is provided. The computer program product includes a computer program stored in a computer readable storage medium. At least one processor of a device reads and executes the computer program from the computer readable storage medium, so that the device performs the method for partition type decision of video coding provided by the embodiments of the present disclosure. BRIEF DESCRIPTION OF DRAWINGS

[0020] FIG. 1 is a flowchart of a method for partition type decision of video coding according to an embodiment of the present disclosure;

[0021] FIG. 2 is a schematic diagram of a coding region corresponding to a coding unit of 32x32 size after performing binary tree partition decision according to an embodiment of the present disclosure;

[0022] FIG. 3 is a schematic diagram of skipping vertical ternary tree partition decision and quad tree partition decision according to an embodiment of the present disclosure;

[0023] FIG. 4 is a flowchart of a method for partition type decision of video coding according to an embodiment of the present disclosure, which includes skipping tree partition decision of other partition directions;

[0024] FIG. 5 is a flowchart of a method for partition type decision of video coding according to an embodiment of the present disclosure, which includes skipping tree partition decision of horizontal partition direction.

[0025] FIG. 6 is a flowchart of a partition type decision method of a video coding containing a tree partition decision process containing a skip containing a vertical partition direction according to an embodiment of the present application;

[0026] FIG. 7 is a flowchart of a partition type decision method of a video coding containing a tree partition decision process containing a skip containing other partition directions according to a preset skip flag according to an embodiment of the present application;

[0027] FIG. 8 is a flowchart of a partition type decision method of a video coding containing a tree partition decision process containing other partition directions of the current coding region according to an embodiment of the present application;

[0028] FIG. 9 is a structural block diagram of a partition type decision device of a video coding according to an embodiment of the present application;

[0029] FIG. 10 is a structural schematic diagram of a partition type decision device of a video coding according to an embodiment of the present application. DETAILED DESCRIPTION

[0030] The embodiments of the present application will be described in further detail below with reference to the drawings and embodiments. It can be understood that the specific embodiments described herein are only used to explain the embodiments of the present application, and not to limit the embodiments of the present application. In addition, it should be noted that, for the convenience of description, only the parts related to the embodiments of the present application are shown in the drawings, not all the structures.

[0031] The terms "first", "second", and the like in the specification and claims of the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be exchanged under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described herein, and the objects distinguished by "first", "second", etc. are generally a class, not limited to the number of objects, for example, the first object can be one or more. In addition, the specification and claims "and / or" means at least one of the connected objects, character " / ", generally indicates that the front and rear associated objects are a "or" relationship.

[0032] The video coding partition type decision method provided in the embodiments of the present application is used to identify the optimal partition direction suitable for the current coding area by using the optimal coding unit partition information after the binary tree partition decision, and to skip the tree partition decision of other partition directions, thereby accelerating the video coding. The specific application scenarios can include video conference, video call, indoor live broadcast, and outdoor live broadcast, etc. The foregoing listed application scenarios are only exemplary and explanatory, and in actual application, the video coding partition type decision method can also be used in video coding in other scenarios, and the embodiments of the present application do not limit this. The present application aims to provide a video coding partition type decision method, device, equipment and storage medium, and solve the problem that the calculation complexity of the video encoder is increased, the coding speed is reduced, and the video coding efficiency is affected due to the need to perform more partition type decisions.

[0033] The execution subject of each step of the video coding partition type decision method provided in the embodiments of the present application can be a computer device, which refers to any electronic device with data calculation, processing and storage capabilities, such as mobile phones, PC (Personal Computer), tablet computers and other terminal devices, and can also be a server or other equipment, and the embodiments of the present application do not limit this.

[0034] FIG. 1 is a flowchart of a video coding partition type decision method provided in the embodiments of the present application, wherein the execution subject of each step in the partition decision method can be a video coding partition type decision device. As shown in FIG. 1, the method includes the following steps:

[0035] In step S101, in the case that the binary tree partition decision is completed in the current coding area, the partition type information corresponding to the first partition depth, the number of coding units corresponding to the second partition depth, and the second optimal partition type of each coding unit are obtained, wherein the partition type information includes the first optimal partition type, and the second partition depth is greater than the first partition depth.

[0036] In the video encoding process, the coding tree unit is taken as the root node, first the non-split type decision is made, then the priority of performing the quadtree splitting or the binary tree splitting is determined according to the value of the split flag. After the binary tree splitting decision of the coding tree unit is completed, it returns to its initial state, and other tree splitting decisions are made, such as vertical ternary tree splitting decision, horizontal ternary tree splitting decision, and quadtree splitting decision, and finally the rate-distortion cost minimum splitting structure is taken as the final splitting result of the coding tree unit. The embodiments of the present application determine whether the other tree splitting decisions can be skipped after the priority of performing the binary tree splitting. The binary tree splitting process can be first horizontal binary tree splitting, then recursive multi-type tree splitting of the splitting result according to the maximum allowed splitting depth; after the splitting is completed, it returns to the previous layer, and continues to perform vertical binary tree splitting, and recursive multi-type tree splitting of the splitting result according to the maximum allowed splitting depth. In the binary tree splitting process, for different splitting depths, the rate-distortion cost minimum splitting type can be taken as the optimal splitting type corresponding to the splitting depth, and the number of coding units obtained by the final splitting of each splitting depth may be different, and the optimal splitting type corresponding to each coding unit may be the same or different. In an embodiment, the first splitting depth can be 1, and in the case of splitting depth 1, the rate-distortion cost calculated by the non-splitting, horizontal binary tree splitting and vertical binary tree splitting respectively can be obtained and compared, and the rate-distortion cost minimum splitting type is taken as the first optimal splitting type. Optionally, the calculation of the rate-distortion cost can comprehensively consider the distortion caused by the current coding mode and the code rate consumed by using the current coding mode for encoding, for example, the related formula for calculating the rate-distortion cost is as follows: J=D+λ mode ·R

[0037] Wherein, D represents the distortion value under the current coding mode, R represents the code rate value under the current coding mode, and λ mode is the Lagrange factor.

[0038] In one embodiment, an optimal partition type flag of a first partition depth can be acquired, and a first optimal partition type corresponding to the first partition depth is determined. In the case that the first partition depth is 1, the optimal partition type flag can be set with three values corresponding to no partition, horizontal binary tree partition and vertical binary tree partition. Since the second partition depth is greater than the first partition depth, in one embodiment, the first partition depth can be 1, and the second partition depth can be 2. If the first optimal partition type corresponding to the first partition depth is vertical binary tree partition or horizontal binary tree partition, the second optimal partition type of the coding unit corresponding to the second partition depth can be horizontal partition or vertical partition, which can be determined according to the comparison result of rate-distortion cost corresponding to different partition types or the optimal partition type flag corresponding to the coding unit. Of course, the values of the first partition depth and the second partition depth are only exemplary described, and the first partition depth and the second partition depth can also refer to partition type information of other value conditions, which is not limited herein. FIG. 2 is a schematic diagram of an encoding area corresponding to a coding unit with a size of 32x32 after performing binary tree partition decision. As shown in FIG. 2, the performed partition type decision of the encoding area 101 includes no partition decision 102, horizontal binary tree partition decision 103 and vertical binary tree partition decision 104. Then, the encoding area 101 needs to return to the original state and continue to perform other tree type partition decisions, which can include horizontal ternary tree partition decision 105, vertical ternary tree partition decision 106 and quadtree partition decision 107. Since the texture information content of the same encoding area is fixed, after the binary tree partition decision, if the partition decision result is that the partition direction of the coding unit in the encoding area is single or most of the coding units correspond to the same partition direction, the probability that the tree type partition corresponding to other partition directions is not suitable for the encoding area is relatively high, and the decision process of the tree type partition can be skipped to achieve the purpose of accelerating encoding. Therefore, the partition information of different partition depths corresponding to the binary tree partition decision can be acquired to identify the optimal partition direction of the current encoding area, and to provide reliable reference information for the partition type decision with low skipping probability.

[0039] In step S102, the number of coding units is compared with a preset number threshold.

[0040] The preset quantity threshold can be a set quantity used to determine whether the number of coding units corresponding to the second division depth effectively represents the texture information of the current coding region. If the number of coding units is greater than or equal to the preset quantity threshold, it can be considered that the texture information of the current coding region is effectively represented, and the optimal division direction of the current coding region can be determined by combining the second optimal division type of each coding unit. If the number of coding units is less than the preset quantity threshold, it can be considered that the texture information of the current coding region cannot be effectively represented, and in addition to considering the second optimal division type of each coding unit, other information needs to be combined for comprehensive judgment of the optimal division direction, for example, rate-distortion cost ratio information of division types corresponding to different division directions at the first division depth. The preset quantity threshold can be adaptively set by a developer according to the possible number of coding units at the second division depth in an actual application scenario, which is not limited herein.

[0041] In step S103, in the case that the first optimal division type is the binary tree division of the preset division direction, the number of coding units is greater than or equal to the preset quantity threshold, and the second optimal division type of each coding unit is the division type of the preset division direction, the tree division decision of other division directions of the current coding region is skipped.

[0042] The preset division direction can include a vertical division direction and a horizontal division direction. By determining the first optimal division type corresponding to the first division depth, a more optimal division direction of the current coding region can be preliminarily judged. By determining that the number of coding units is greater than or equal to a preset number threshold and the second optimal division type of each coding unit is a division type of the preset division direction, it can be further determined whether the texture information of the current coding region matches the obtained more optimal division direction, so as to determine whether to skip the tree type division decision of other division directions of the current coding region. In an embodiment, the tree type division decision to be executed can be a ternary tree division decision and a quad tree division decision. If the first optimal division type is a horizontal binary tree division, the number of coding units is greater than or equal to the preset number threshold, and the second optimal division type of each coding unit is horizontal division, the horizontal division direction can be regarded as the more optimal division direction corresponding to the current coding region. Therefore, the tree type division decision including the vertical division direction, such as the vertical ternary tree division decision and the quad tree division decision, can be skipped. In an embodiment, FIG. 3 is a schematic diagram of skipping the vertical ternary tree division decision and the quad tree division decision based on the video coding division type decision method provided by the embodiment of the present application. As shown in FIG. 3, the executed division type decisions of the coding region 201 include the no division decision 202, the horizontal binary tree division decision 203, and the vertical binary tree division decision 204. Then, the coding region 201 needs to return to the original state and continue to perform other tree type division decisions to be executed, which can include the horizontal ternary tree division decision 205, the vertical ternary tree division decision 206, and the quad tree division decision 207. If the first optimal division type corresponding to the first division depth is the vertical binary tree division decision 204, the possible values of the number of coding units corresponding to the second division depth are 2, 3, and 4. In order to effectively combine more delicate texture information, the preset number threshold can be set to 3. Therefore, in combination with the condition that the number of coding units is greater than or equal to 3 and the second optimal division type of each coding unit is vertical division, the first division result 208 and the second division result 209 corresponding to the second division depth in FIG. 3 can regard the vertical division direction as the more optimal division direction corresponding to the current coding region. Therefore, the horizontal ternary tree division decision 205 and the quad tree division decision 207 including the horizontal division direction can be skipped.

[0043] In the foregoing, in a case where the binary tree partition decision is completed in the current coding region, the partition type information corresponding to the first partition depth, the number of coding units corresponding to the second partition depth, and the second optimal partition type of each coding unit are obtained, wherein the partition type information comprises the first optimal partition type, and the second partition depth is greater than the first partition depth; then, the number of coding units is compared with a preset number threshold; and finally, in a case where the first optimal partition type is the binary tree partition in the preset partition direction, the number of coding units is greater than or equal to the preset number threshold, and the second optimal partition type of each coding unit is the partition type in the preset partition direction, the tree partition decision in the current coding region in other partition directions is skipped. In the foregoing scheme, by obtaining the partition type information corresponding to different partition depths, the optimal coding unit partition information after the binary tree partition decision can be effectively referenced, and reliable judgment basis for determining whether to skip the other tree partition decision is provided; according to the comparison result of the number of coding units and the preset number threshold and the optimal partition type statistics in different partition depths, the texture information of the current coding region can be reasonably utilized, the best partition direction can be effectively identified, it is determined whether the tree partition decision in the current coding region in other partition directions is executed, the partition type decision with a smaller optimal probability is skipped on the premise of less loss of coding quality, the calculation complexity of the video encoder is reduced, the coding speed is improved, and the video coding efficiency is improved. Furthermore, the use rate of the computing resources of the server side is improved, the server cost is saved, the CPU use rate of the client side is reduced, and the user experience is further improved.

[0044] FIG. 4 is a flowchart of a partition type decision method provided by an embodiment of the application and comprising a tree partition decision process in which a tree partition decision in a current coding region in other partition directions is skipped. As shown in FIG. 4, the method comprises the following steps:

[0045] In a case where the binary tree partition decision is completed in the current coding region, the partition type information corresponding to the first partition depth, the number of coding units corresponding to the second partition depth, and the second optimal partition type of each coding unit are obtained, wherein the partition type information comprises the first optimal partition type, and the second partition depth is greater than the first partition depth.

[0046] In a case where the binary tree partition decision is completed in the current coding region, the partition type information corresponding to the first partition depth, the number of coding units corresponding to the second partition depth, and the second optimal partition type of each coding unit are obtained, wherein the partition type information comprises the first optimal partition type, and the second partition depth is greater than the first partition depth.

[0047] In a case where the first optimal partition type is the vertical binary tree partition, the number of coding units is greater than or equal to the preset number threshold, and the second optimal partition type of each coding unit is the vertical partition, the horizontal ternary tree partition decision and the quadtree partition decision in the current coding region are skipped.

[0048] The tree partition decision to be performed on the current coding region includes a ternary tree partition decision and a quadtree partition decision. If the first optimal partition type is a vertical binary tree partition, it can be preliminarily determined that the optimal partition direction of the current coding region is a vertical partition direction. Then, in combination with the determination result that the number of coding units is greater than or equal to a preset number threshold and the second optimal partition type of each coding unit is a vertical partition, it can be determined that the probability that the texture information of the current coding region is of a vertical type is relatively high. Accordingly, the probability of performing a tree partition decision including a horizontal partition direction is relatively small, and the ternary tree partition decision and the quadtree partition decision in the horizontal direction can be skipped.

[0049] In one embodiment, the specific process of skipping the ternary tree partition decision and the quadtree partition decision in the horizontal direction of the current coding region can be as follows:

[0050] A first skip flag corresponding to the ternary tree partition decision and the quadtree partition decision in the horizontal direction of the current coding region is set to a true value. Before performing the ternary tree partition decision and the quadtree partition decision in the horizontal direction of the current coding region, if it is detected that the first skip flag is the true value, the ternary tree partition decision and the quadtree partition decision in the horizontal direction are skipped.

[0051] The ternary tree partition decision and the quadtree partition decision in the horizontal direction of the current coding region can be pre-provided with a first skip flag. If the first skip flag is a true value, the ternary tree partition decision and the quadtree partition decision in the horizontal direction are skipped. If the first skip flag is a false value, the ternary tree partition decision and the quadtree partition decision in the horizontal direction are not skipped. In the case where the first optimal partition type is a vertical binary tree partition, the number of coding units is greater than or equal to a preset number threshold, and the second optimal partition type of each coding unit is a vertical partition, the first skip flag of the current coding region can be set to a true value. Therefore, before the ternary tree partition decision and the quadtree partition decision in the horizontal direction of the current coding region are performed, whether to skip the ternary tree partition decision and the quadtree partition decision in the horizontal direction can be determined by detecting the first skip flag.

[0052] In one embodiment, the horizontal ternary tree partitioning decision and the quadtree partitioning decision corresponding to the current coding region can be provided with a first execution flag. If the first execution flag is true, the horizontal ternary tree partitioning decision and the quadtree partitioning decision are executed. If the first execution flag is false, the horizontal ternary tree partitioning decision and the quadtree partitioning decision are skipped. In the case where the first optimal partition type is vertical binary tree partitioning, the number of coding units is greater than or equal to a preset number threshold, and the second optimal partition type of each coding unit is vertical partitioning, the first execution flag corresponding to the current coding region can be set to false. Therefore, before the horizontal ternary tree partitioning decision and the quadtree partitioning decision of the current coding region are needed to be performed, it can be determined whether to execute the horizontal ternary tree partitioning decision and the quadtree partitioning decision by detecting the first execution flag.

[0053] In step S204, in the case where the first optimal partition type is horizontal binary tree partitioning, the number of coding units is greater than or equal to a preset number threshold, and the second optimal partition type of each coding unit is horizontal partitioning, the vertical ternary tree partitioning decision and the quadtree partitioning decision of the current coding region are skipped.

[0054] In one embodiment, in the case where the first optimal partition type is horizontal binary tree partitioning, it can be initially determined that the optimal partition direction of the current coding region is the horizontal partitioning direction. Then, in combination with the determination that the number of coding units is greater than or equal to a preset number threshold and the second optimal partition type of each coding unit is horizontal partitioning, it can be determined that the probability that the texture information of the current coding region is of the horizontal type is relatively high. Accordingly, the probability that the tree type partitioning decision containing the vertical partitioning direction is executed is relatively small, and the vertical ternary tree partitioning decision and the quadtree partitioning decision can be skipped.

[0055] In one embodiment, the specific process of skipping the vertical ternary tree partitioning decision and the quadtree partitioning decision of the current coding region can be as follows:

[0056] A second skip flag corresponding to the vertical ternary tree partitioning decision and the quadtree partitioning decision of the current coding region is set to true. Before the vertical ternary tree partitioning decision and the quadtree partitioning decision of the current coding region are performed, if it is detected that the second skip flag is true, the vertical ternary tree partitioning decision and the quadtree partitioning decision are skipped.

[0057] The vertical ternary tree partition decision and the quadtree partition decision of the current coding region can be provided with a second skip flag. If the second skip flag is true, the vertical ternary tree partition decision and the quadtree partition decision are skipped. If the second skip flag is false, the vertical ternary tree partition decision and the quadtree partition decision are not skipped. In the case that the first optimal partition type is vertical binary tree partition, the number of coding units is greater than or equal to a preset number threshold, and the second optimal partition type of each coding unit is vertical partition, the second skip flag of the current coding region can be set to true. Therefore, before the vertical ternary tree partition decision and the quadtree partition decision of the current coding region are needed, it can be determined whether to skip the vertical ternary tree partition decision and the quadtree partition decision by detecting the second skip flag.

[0058] In an embodiment, the vertical ternary tree partition decision and the quadtree partition decision corresponding to the current coding region can be provided with a second execution flag. If the second execution flag is true, the vertical ternary tree partition decision and the quadtree partition decision are executed. If the second execution flag is false, the vertical ternary tree partition decision and the quadtree partition decision are skipped. In the case that the first optimal partition type is vertical binary tree partition, the number of coding units is greater than or equal to a preset number threshold, and the second optimal partition type of each coding unit is vertical partition, the second execution flag corresponding to the current coding region can be set to false. Therefore, before the vertical ternary tree partition decision and the quadtree partition decision of the current coding region are needed, it can be determined whether to execute the vertical ternary tree partition decision and the quadtree partition decision by detecting the second execution flag.

[0059] By judging the partition type information corresponding to the first partition depth and the second partition depth, the optimal partition direction with higher probability corresponding to the current coding region can be identified in combination with the optimal partition result of the binary tree partition, which is beneficial to accurately determine the tree partition decision of other partition directions that need to be skipped.

[0060] FIG. 5 is a flowchart of a partition type decision method provided by an embodiment of the present application, which includes a tree partition decision process of skipping a partition direction including horizontal partition. As shown in FIG. 5, the method includes the following steps:

[0061] In step S301, in the case that the binary tree partition decision of the current coding region is completed, the partition type information corresponding to the first partition depth, the number of coding units corresponding to the second partition depth, and the second optimal partition type of each coding unit are obtained. The partition type information includes the first optimal partition type, the first rate-distortion cost corresponding to vertical binary tree partition, and the second rate-distortion cost corresponding to horizontal binary tree partition. The second partition depth is greater than the first partition depth.

[0062] Step S302, compare the number of coding units with the preset number threshold.

[0063] Step S303, in the case that the first optimal partition type is vertical binary tree partition, the number of coding units is greater than or equal to the preset number threshold, and the second optimal partition type of each coding unit is vertical partition, skip the horizontal ternary tree partition decision and the quadtree partition decision of the current coding region.

[0064] Step S304, in the case that the first optimal partition type is horizontal binary tree partition, the number of coding units is greater than or equal to the preset number threshold, and the second optimal partition type of each coding unit is horizontal partition, skip the vertical ternary tree partition decision and the quadtree partition decision of the current coding region.

[0065] Step S305, in the case that the first optimal partition type is vertical binary tree partition, the number of coding units is less than the preset number threshold, the first rate-distortion cost is less than the first preset proportion of the second rate-distortion cost, and the second optimal partition type of each coding unit is vertical partition, skip the horizontal ternary tree partition decision and the quadtree partition decision of the current coding region.

[0066] In the case that the first optimal partition type is vertical binary tree partition, it can be initially determined that the more optimal partition direction of the current coding region is the vertical partition direction, but since the number of coding units is less than the preset number threshold, the texture information of the current coding region cannot be effectively represented, so in addition to referring to the second optimal partition type of each coding unit, the comparison result of the rate-distortion cost corresponding to the vertical binary tree partition and the horizontal binary tree partition also needs to be referred to. In order to ensure that the vertical partition direction is the optimal partition direction corresponding to the current coding region with a high probability, in addition to determining that the second optimal partition type of each coding unit is vertical partition, it is also necessary to determine that the first rate-distortion cost is less than the first preset proportion of the second rate-distortion cost. The first preset proportion can be 0.4 or 0.5, etc., to improve the possibility that the vertical partition direction is the optimal partition direction corresponding to the current coding region.

[0067] For the case that the first optimal partition type is vertical binary tree partition and the second optimal partition type of each coding unit is vertical partition, but the number of coding units is less than the preset number threshold, the possibility that the vertical binary tree partition is more matched with the current coding region than the horizontal binary tree partition, and the vertical partition direction is the optimal partition direction corresponding to the current coding region can be obtained by determining that the first rate-distortion cost is less than the first preset proportion of the second rate-distortion cost, so as to ensure that the coding quality loss caused by skipping the horizontal ternary tree partition decision and the quadtree partition decision of the current coding region is small.

[0068] FIG. 6 is a flowchart of a partition type decision method including a tree partition decision process of skipping a vertical partition direction according to an embodiment of the present application. As shown in FIG. 6, the method includes the following steps.

[0069] In step S401, in a case where the binary tree partition decision of the current coding region is completed, the partition type information corresponding to the first partition depth, the number of coding units corresponding to the second partition depth, and the second optimal partition type of each coding unit are obtained, wherein the partition type information includes the first optimal partition type, the first rate-distortion cost corresponding to the vertical binary tree partition, and the second rate-distortion cost corresponding to the horizontal binary tree partition, and the second partition depth is greater than the first partition depth.

[0070] In step S402, the number of coding units is compared with a preset number threshold.

[0071] In step S403, in a case where the first optimal partition type is the vertical binary tree partition, the number of coding units is greater than or equal to the preset number threshold, and the second optimal partition type of each coding unit is the vertical partition, the horizontal ternary tree partition decision and the quadtree partition decision of the current coding region are skipped.

[0072] In step S404, in a case where the first optimal partition type is the horizontal binary tree partition, the number of coding units is greater than or equal to the preset number threshold, and the second optimal partition type of each coding unit is the horizontal partition, the vertical ternary tree partition decision and the quadtree partition decision of the current coding region are skipped.

[0073] In step S405, in a case where the first optimal partition type is the vertical binary tree partition, the number of coding units is less than the preset number threshold, the first rate-distortion cost is less than a first preset proportion of the second rate-distortion cost, and the second optimal partition type of each coding unit is the vertical partition, the horizontal ternary tree partition decision and the quadtree partition decision of the current coding region are skipped.

[0074] In step S406, in a case where the first optimal partition type is the horizontal binary tree partition, the number of coding units is less than the preset number threshold, the second rate-distortion cost is less than a second preset proportion of the first rate-distortion cost, and the second optimal partition type of each coding unit is the horizontal partition, the vertical ternary tree partition decision and the quadtree partition decision of the current coding region are skipped.

[0075] If the first optimal partition type is horizontal binary tree partition, it can be preliminarily judged that the optimal partition direction of the current coding region is the horizontal partition direction. However, since the number of coding units is less than the preset number threshold, the texture information of the current coding region cannot be effectively represented. Therefore, in addition to referring to the second optimal partition type of each coding unit, the comparison result of the rate-distortion cost corresponding to the vertical binary tree partition and the horizontal binary tree partition also needs to be referred to. In order to ensure that the horizontal partition direction is the optimal partition direction corresponding to the current coding region with a high probability, in addition to determining that the second optimal partition type of each coding unit is horizontal partition, it also needs to be determined that the second rate-distortion cost is less than the first rate-distortion cost by a second preset proportion. The second preset proportion can be 0.4 or 0.5, etc., so as to improve the possibility that the horizontal partition direction is the optimal partition direction corresponding to the current coding region.

[0076] For the case that the first optimal partition type is horizontal binary tree partition and the second optimal partition type of each coding unit is horizontal partition, but the number of coding units is less than the preset number threshold, it can be obtained by determining that the second rate-distortion cost is less than the first rate-distortion cost by a second preset proportion that the horizontal binary tree partition has a higher matching degree with the current coding region than the vertical binary tree partition, the possibility that the horizontal partition direction is the optimal partition direction corresponding to the current coding region is greater, and the coding quality loss caused by skipping the vertical three-triangular tree partition decision and the four-quadrangular tree partition decision of the current coding region is smaller.

[0077] FIG. 7 is a flowchart of a partition type decision method provided by an embodiment of the present application, which includes a tree type partition decision process of skipping other partition directions according to a preset skip flag. As shown in FIG. 7, the method includes the following steps:

[0078] In step S501, in the case that the binary tree partition decision of the current coding region is completed, the partition type information corresponding to the first partition depth, the number of coding units corresponding to the second partition depth, and the second optimal partition type of each coding unit are obtained. The partition type information includes the first optimal partition type, and the second partition depth is greater than the first partition depth.

[0079] In step S502, the number of coding units is compared with the preset number threshold.

[0080] In a case where the first optimal partition type is the binary tree partition of the preset partition direction, the number of coding units is greater than or equal to the preset number threshold, and the second optimal partition type of each coding unit is the partition type of the preset partition direction, the preset skip flag corresponding to the tree type partition decision of the other partition direction of the current coding region is set to a true value, and in a case where the preset skip flag is detected to be the true value before the tree type partition decision of the other partition direction of the current coding region is performed, the tree type partition decision of the other partition direction is skipped.

[0081] In a case where the first optimal partition type is the binary tree partition of the preset partition direction, the number of coding units is greater than or equal to the preset number threshold, and the second optimal partition type of each coding unit is the partition type of the preset partition direction, the preset skip flag corresponding to the tree type partition decision of the other partition direction of the current coding region is set to a true value, and in a case where the preset skip flag is detected to be the true value before the tree type partition decision of the other partition direction of the current coding region is performed, the tree type partition decision of the other partition direction is skipped.

[0082] In a case where the first optimal partition type is the binary tree partition of the preset partition direction, the number of coding units is greater than or equal to the preset number threshold, and the second optimal partition type of each coding unit is the partition type of the preset partition direction, the preset skip flag corresponding to the tree type partition decision of the other partition direction of the current coding region is set to a true value, and in a case where the preset skip flag is detected to be the true value before the tree type partition decision of the other partition direction of the current coding region is performed, the tree type partition decision of the other partition direction is skipped.

[0083] In a case where the first optimal partition type is the binary tree partition of the preset partition direction, the number of coding units is greater than or equal to the preset number threshold, and the second optimal partition type of each coding unit is the partition type of the preset partition direction, the preset skip flag corresponding to the tree type partition decision of the other partition direction of the current coding region is set to a true value, and in a case where the preset skip flag is detected to be the true value before the tree type partition decision of the other partition direction of the current coding region is performed, the tree type partition decision of the other partition direction is skipped.

[0084] FIG. 8 is a flowchart of a partition type decision method provided by an embodiment of the present application, which includes a process of performing a tree type partition decision of the other partition direction of a current coding region. As shown in FIG. 8, the method includes the following steps.

[0085] In a case where the binary tree partition decision of the current coding region is completed, the partition type information corresponding to the first partition depth, the number of coding units corresponding to the second partition depth, and the second optimal partition type of each coding unit are obtained, wherein the partition type information comprises the first optimal partition type, and the second partition depth is greater than the first partition depth.

[0086] In a case where the binary tree partition decision of the current coding region is completed, the partition type information corresponding to the first partition depth, the number of coding units corresponding to the second partition depth, and the second optimal partition type of each coding unit are obtained, wherein the partition type information comprises the first optimal partition type, and the second partition depth is greater than the first partition depth.

[0087] In a case where the binary tree partition decision of the current coding region is completed, the partition type information corresponding to the first partition depth, the number of coding units corresponding to the second partition depth, and the second optimal partition type of each coding unit are obtained, wherein the partition type information comprises the first optimal partition type, and the second partition depth is greater than the first partition depth.

[0088] In a case where the binary tree partition decision of the current coding region is completed, the partition type information corresponding to the first partition depth, the number of coding units corresponding to the second partition depth, and the second optimal partition type of each coding unit are obtained, wherein the partition type information comprises the first optimal partition type, and the second partition depth is greater than the first partition depth.

[0089] In a case where the binary tree partition decision of the current coding region is completed, the partition type information corresponding to the first partition depth, the number of coding units corresponding to the second partition depth, and the second optimal partition type of each coding unit are obtained, wherein the partition type information comprises the first optimal partition type, and the second partition depth is greater than the first partition depth.

[0090] In a case where the binary tree partition decision of the current coding region is completed, the partition type information corresponding to the first partition depth, the number of coding units corresponding to the second partition depth, and the second optimal partition type of each coding unit are obtained, wherein the partition type information comprises the first optimal partition type, and the second partition depth is greater than the first partition depth.

[0091] FIG. 9 is a structural block diagram of a partition type decision device for video coding provided by an embodiment of the present application. The device is configured to execute the partition type decision method for video coding provided by the above-described embodiments, and has the function modules and beneficial effects corresponding to the execution method. As shown in FIG. 9, the device specifically comprises:

[0092] The partition information obtaining module 301 is configured to, in a case where the binary tree partition decision of the current coding region is completed, obtain the partition type information corresponding to the first partition depth, the number of coding units corresponding to the second partition depth, and the second optimal partition type of each coding unit, wherein the partition type information comprises the first optimal partition type, and the second partition depth is greater than the first partition depth.

[0093] The unit quantity comparison module 302 is configured to compare the coding unit quantity with a preset quantity threshold.

[0094] The partition decision skipping module 303 is configured to skip the tree partition decision of other partition directions of the current coding region in the case that the first optimal partition type is the binary tree partition of the preset partition direction, the coding unit quantity is greater than or equal to the preset quantity threshold, and the second optimal partition type of each coding unit is the partition type of the preset partition direction.

[0095] In the above, in the case that the binary tree partition decision of the current coding region is completed, the partition type information corresponding to the first partition depth, the coding unit quantity corresponding to the second partition depth, and the second optimal partition type of each coding unit are obtained, wherein the partition type information includes the first optimal partition type, and the second partition depth is greater than the first partition depth; then, the coding unit quantity is compared with the preset quantity threshold; finally, in the case that the first optimal partition type is the binary tree partition of the preset partition direction, the coding unit quantity is greater than or equal to the preset quantity threshold, and the second optimal partition type of each coding unit is the partition type of the preset partition direction, the tree partition decision of other partition directions of the current coding region is skipped. In the above scheme, by obtaining the partition type information corresponding to different partition depths, the optimal coding unit partition information after the binary tree partition decision can be effectively referred to, and reliable judgment basis for determining whether to skip the tree partition decision of other partition directions is provided; according to the comparison result of the coding unit quantity and the preset quantity threshold and the optimal partition type statistics of different partition depths, the texture information of the current coding region can be reasonably utilized, the best partition direction can be effectively identified, it is determined whether to execute the tree partition decision of other partition directions of the current coding region, the partition type decision to be partitioned with a smaller optimal probability is skipped on the premise of less loss of coding quality, the calculation complexity of the video encoder is reduced, the encoding speed is improved, and the video coding efficiency is improved. Further, the use rate of the computing resources of the server side is improved, the server cost is saved, the CPU use rate of the client side is reduced, and the user experience is further improved.

[0096] In one possible embodiment, the partition decision skipping module 303 is configured to:

[0097] In the case that the first optimal partition type is the vertical binary tree partition, the coding unit quantity is greater than or equal to the preset quantity threshold, and the second optimal partition type of each coding unit is the vertical partition, the horizontal ternary tree partition decision and the quadtree partition decision of the current coding region are skipped.

[0098] In a case where the first optimal partition type is horizontal binary tree partition, the number of coding units is greater than or equal to the preset number threshold, the second optimal partition type of each coding unit is horizontal partition, and the second rate-distortion cost is less than the first rate-distortion cost, the vertical ternary tree partition decision and the quadtree partition decision of the current coding region are skipped.

[0099] In one possible embodiment, the first skip decision module is further configured to:

[0100] In a case where the first optimal partition type is vertical binary tree partition, the number of coding units is less than the preset number threshold, the first rate-distortion cost is less than the second rate-distortion cost by the first preset proportion, and the second optimal partition type of each coding unit is vertical partition, the horizontal ternary tree partition decision and the quadtree partition decision of the current coding region are skipped.

[0101] In one possible embodiment, the second skip decision module is further configured to:

[0102] In a case where the first optimal partition type is horizontal binary tree partition, the number of coding units is less than the preset number threshold, the second rate-distortion cost is less than the first rate-distortion cost by the second preset proportion, and the second optimal partition type of each coding unit is horizontal partition, the vertical ternary tree partition decision and the quadtree partition decision of the current coding region are skipped.

[0103] In one possible embodiment, the partition decision skip module 303 is further configured to:

[0104] The preset skip flag corresponding to the tree type partition decision of the current coding region in the other partition direction is set to a true value, and in a case where the preset skip flag is detected to be the true value before the tree type partition decision in the other partition direction of the current coding region is performed, the tree type partition decision in the other partition direction is skipped.

[0105] In one possible embodiment, the first skip decision module is further configured to:

[0106] The first skip flag corresponding to the horizontal ternary tree partition decision and the quadtree partition decision of the current coding region is set to a true value, and in a case where the first skip flag is detected to be the true value before the horizontal ternary tree partition decision and the quadtree partition decision of the current coding region are performed, the horizontal ternary tree partition decision and the quadtree partition decision are skipped.

[0107] In one possible embodiment, the partition decision execution module is further configured to:

[0108] In a case where the first optimal partition type is binary tree partition in the preset partition direction, and there is a case where the second optimal partition type of the coding unit is a partition type in the other partition direction, the tree type partition decision in the other partition direction of the current coding region is executed.

[0109] FIG. 10 is a structural schematic diagram of a video coding partition type decision device provided by an embodiment of the present application. As shown in FIG. 10, the device includes a processor 401, a memory 402, an input device 403 and an output device 404. The number of processors 401 in the device can be one or more, and one processor 401 is taken as an example in FIG. 10. The processor 401, the memory 402, the input device 403 and the output device 404 in the device can be connected through a bus or other means, and the connection through a bus is taken as an example in FIG. 10. The memory 402 is a computer readable storage medium, which can be configured to store software programs, computer executable programs and modules, such as program instructions / modules of the video coding partition type decision method in the embodiments of the present application. The processor 401 executes the software programs, instructions and modules stored in the memory 402, so as to perform various function applications and data processing of the device, that is, to implement the video coding partition type decision method described above. The input device 403 can be configured to receive input digital or character information, and to generate key signal input related to user settings and function control of the device. The output device 404 can include a display device such as a display screen.

[0110] The embodiments of the present application also provide a nonvolatile storage medium containing computer executable instructions, which are configured to perform a video coding partition type decision method described in the above embodiments when executed by a computer processor, and the method includes: in the case that a binary tree partition decision of a current coding region is completed, obtaining partition type information corresponding to a first partition depth, a number of coding units corresponding to a second partition depth and a second optimal partition type of each coding unit, wherein the partition type information includes a first optimal partition type, the second partition depth is greater than the first partition depth, and the number of coding units is compared with a preset number threshold; in the case that the first optimal partition type is a binary tree partition of a preset partition direction, the number of coding units is greater than or equal to the preset number threshold, and the second optimal partition type of each coding unit is a partition type of a preset partition direction, the tree type partition decision of the current coding region containing other partition directions is skipped.

[0111] It is worth noting that in the embodiments of the video coding partition type decision device described above, each unit and module is only divided according to the function logic, but is not limited to the above division, as long as the corresponding function can be realized; in addition, the specific name of each functional unit is only for convenient distinction, and does not configure to limit the protection scope of the embodiments of the present application.

[0112] In some possible implementation, each of the aspects of the method provided by the present application can also be implemented as a program product in the form of a computer program product, which includes program codes configured to cause a computer device to perform the steps of the method according to various exemplary embodiments of the present application described above in the specification when the program product is run on the computer device, for example, the computer device can perform the video coding partition type decision method described in the embodiments of the present application. The program product can be implemented by any combination of one or more computer readable media.

Claims

1. A method for classifying video coding types, wherein, include: When a binary tree partitioning decision is made in the current coding region, the partitioning type information corresponding to the first partitioning depth, the number of coding units corresponding to the second partitioning depth, and the second optimal partitioning type for each coding unit are obtained. The partitioning type information includes the first optimal partitioning type, and the second partitioning depth is greater than the first partitioning depth. The number of coding units is compared with a preset number threshold; If the first optimal partitioning type is a binary tree partitioning with a preset partitioning direction, the number of coding units is greater than or equal to the preset number threshold, and the second optimal partitioning type of each coding unit is a partitioning type with a preset partitioning direction, then skip the tree partitioning decision that includes other partitioning directions for the current coding region.

2. The video coding classification decision method according to claim 1, wherein, The step of skipping tree partitioning decisions involving other partitioning directions in the current encoding region when the first optimal partitioning type is a binary tree partition with a preset partitioning direction, the number of encoding units is greater than or equal to the preset number threshold, and the second optimal partitioning type of each encoding unit is a partitioning type with a preset partitioning direction, includes: If the first optimal partitioning type is vertical binary tree partitioning, the number of coding units is greater than or equal to the preset number threshold, and the second optimal partitioning type of each coding unit is vertical partitioning, skip the horizontal ternary tree partitioning decision and the quadtree partitioning decision of the current coding region. If the first optimal partitioning type is a horizontal binary tree partitioning, the number of coding units is greater than or equal to the preset number threshold, and the second optimal partitioning type of each coding unit is a horizontal partitioning, then skip the vertical ternary tree partitioning decision and the quadtree partitioning decision of the current coding region.

3. The video coding classification decision method according to claim 2, wherein, The partitioning type information also includes the first rate-distortion cost corresponding to vertical binary tree partitioning and the second rate-distortion cost corresponding to horizontal binary tree partitioning. The method further includes: When the first optimal partitioning type is vertical binary tree partitioning, the number of coding units is less than the preset number threshold, the first rate distortion cost is less than the first preset ratio of the second rate distortion cost, and the second optimal partitioning type of each coding unit is vertical partitioning, skip the horizontal ternary tree partitioning decision and the quadtree partitioning decision of the current coding region.

4. The video coding classification decision method according to claim 3, wherein, The method further includes: When the first optimal partitioning type is a horizontal binary tree partitioning, the number of coding units is less than the preset number threshold, the second rate distortion cost is less than the second preset ratio of the first rate distortion cost, and the second optimal partitioning type of each coding unit is a horizontal partitioning, the vertical ternary tree partitioning decision and the quadtree partitioning decision of the current coding region are skipped.

5. The video coding classification decision method according to claim 1 or 2, wherein, The tree partitioning decision that skips the current encoded region and includes other partitioning directions includes: Set the preset skip flag corresponding to the tree partitioning decision containing other partitioning directions in the current encoding region to true. If the preset skip flag is detected to be true before making the tree partitioning decision containing other partitioning directions in the current encoding region, skip the tree partitioning decision containing other partitioning directions.

6. The video coding classification decision method according to claim 2, wherein, The decision to skip the horizontal ternary tree partitioning and quadtree partitioning of the current encoding region includes: Set the first skip flag corresponding to the horizontal ternary tree partitioning decision and the quadtree partitioning decision of the current encoding region to true. If the first skip flag is detected to be true before making the horizontal ternary tree partitioning decision and the quadtree partitioning decision of the current encoding region, skip the horizontal ternary tree partitioning decision and the quadtree partitioning decision.

7. The video coding classification decision method according to any one of claims 1-6, wherein, The method further includes: If the first optimal partitioning type is a binary tree partitioning with a preset partitioning direction, and there exists a second optimal partitioning type for an encoding unit that is a partitioning type with other partitioning directions, then the tree partitioning decision for the current encoding region that includes other partitioning directions is executed.

8. A video coding classification decision device, wherein, include: The partitioning information acquisition module is configured to acquire, when a binary tree partitioning decision is completed in the current coding region, partitioning type information corresponding to a first partitioning depth, number of coding units corresponding to a second partitioning depth, and second optimal partitioning type for each coding unit, wherein the partitioning type information includes a first optimal partitioning type, and the second partitioning depth is greater than the first partitioning depth. The unit quantity comparison module is configured to compare the number of the coding units with a preset quantity threshold; The partitioning decision skipping module is configured to skip tree partitioning decisions that include other partitioning directions in the current encoding region when the first optimal partitioning type is a binary tree partitioning with a preset partitioning direction, the number of encoding units is greater than or equal to the preset number threshold, and the second optimal partitioning type of each encoding unit is a partitioning type with a preset partitioning direction.

9. A video coding classification decision device, the device comprising: One or more processors; A storage device configured to store one or more programs, which, when executed by one or more processors, cause the one or more processors to implement the video coding partitioning type decision method according to any one of claims 1-7.

10. A non-volatile storage medium storing computer-executable instructions, which, when executed by a computer processor, are configured to perform the video coding classification decision method according to any one of claims 1-7.

11. A computer program product comprising a computer program, wherein, When the computer program is executed by the processor, it implements the video coding classification decision method according to any one of claims 1-7.

Images