Point cloud geometric encoding method, decoding method, encoding device, and decoding device

Abstract

The present invention discloses a point cloud geometric encoding method, a decoding method, an encoding device and a decoding device. The point cloud is defined in a tree structure, and each node in the tree structure includes a plurality of child nodes. The point cloud geometric decoding method includes the steps of: determining the context of the current child node according to the combined occupancy number of adjacent nodes of the current child node; performing entropy decoding on the point cloud geometric code stream according to the context of the current child node to obtain the occupancy information of the current child node. According to the present invention, the context of the current child node is determined according to the combined occupancy number of adjacent nodes of the current child node. The combined occupancy number takes into account the occupancy numbers of adjacent nodes of multiple types of adjacent relationships of the current child node, and determines the context based on the combined occupancy number according to the geometric structure of the adjacent nodes, better utilizing the spatial distribution correlation of adjacent nodes and improving the point cloud geometric compression performance.

Description

Technical Field

[0001] The present invention relates to the technical field of point cloud processing, and particularly to an octree-based point cloud geometric encoding method, a decoding method, an encoding device, and a decoding device. Background Art

[0002] Point cloud compression is mainly divided into geometric compression and attribute compression. Currently, in the geometric compression framework described in the test platform TMC13v12 (Test Model for Category 1&3 version 12) provided by the international standard organization (Moving Picture Experts Group, MPEG), the encoder gradually divides the (aligned bounding box) point cloud into eight child nodes. Only non-empty sub-voxels continue to be subdivided. According to the data structure characteristics of the octree, the position of each voxel is represented by its unit center.

[0003] There is also a point cloud geometric compression method described in the test platform PCRM v9.0 provided by the current Chinese AVS (Audio Video coding Standard) point cloud compression working group, which mainly constructs a hybrid traversal octree, generates a space occupancy code, and encodes the occupancy code through context.

[0004] In octree-based geometric encoding, at each octree node division, the space occupancy code of the node contains eight bits (b0b1b2b3b4b5b6b7), respectively representing the occupancy of the eight child nodes of the node. Entropy encoding is performed on the occupancy bit code of each child node using a separate context, and the context adaptive entropy encoder (CABAC) widely used in the AVS2 standard is adopted to encode each bit (bit or bin) of the space occupancy code in order to achieve a better compression effect. The encoding is mainly divided into two parts: (1) context selection; (2) binary arithmetic encoding.

[0005] In the existing octree geometric encoding in the AVS point cloud encoding standard, first, the occupancy number C1 of 3 coplanar adjacent nodes of the current child node is considered, and then the occupancy number C2 of 3 collinear adjacent nodes is considered. These occupancy numbers only respectively count the number of coplanar adjacent nodes and the number of collinear adjacent nodes, without considering the geometric structure of the coplanar and collinear adjacent nodes, and cannot make good use of the spatial correlation of adjacent nodes, resulting in not selecting a suitable probability and reducing the compression performance.

[0006] Therefore, the existing technology still needs to be improved and developed. Summary of the Invention

[0007] The present invention provides a point cloud geometric encoding method, a decoding method, an encoding device, and a decoding device. According to the combined occupancy of adjacent nodes of the current child node of the octree and considering the geometric structure of the adjacent nodes, the context of the current child node is determined.

[0008] The technical solution adopted by the present invention is as follows:

[0009] A point cloud geometric decoding method, wherein the point cloud is defined in a tree structure, and each node in the tree structure includes a plurality of child nodes. The method is characterized by comprising the following steps:

[0010] Determine the context of the current child node according to the combined occupancy of adjacent nodes of the current child node;

[0011] Entropy-decode the point cloud geometric code stream according to the context of the current child node to obtain the occupancy information of the current child node.

[0012] The point cloud geometric decoding method is characterized in that the step of determining the context of the current child node according to the combined occupancy of adjacent nodes of the current child node includes:

[0013] Determine the context of the current child node according to the combined occupancy of coplanar and collinear adjacent nodes of the current child node and in accordance with the geometric structure of the adjacent nodes;

[0014] Or, determine the context of the current child node according to the combined occupancy of coplanar, collinear, and concurrent adjacent nodes of the current child node and in accordance with the geometric structure of the adjacent nodes.

[0015] The point cloud geometric decoding method is characterized in that the step of determining the context of the current child node according to the combined occupancy of adjacent nodes of the current child node and in accordance with the geometric structure of the adjacent nodes includes:

[0016] Determine the context of the current child node according to the combined occupancy of adjacent nodes of the current child node and in accordance with all geometric structure types of the adjacent nodes under the combined occupancy;

[0017] Or, determine the context of the current child node according to the combined occupancy of adjacent nodes of the current child node, in accordance with all geometric structure types of the adjacent nodes under the combined occupancy, and reduce the geometric structure types of the adjacent nodes according to the number of coplanar adjacent nodes.

[0018] The point cloud geometric decoding method is characterized in that determining the context of the current child node according to the combined occupancy of adjacent nodes of the current child node includes:

[0019] Determine the context based on the adjacent nodes of the child node according to the combined occupancy of the adjacent nodes of the current child node;

[0020] Determine the context based on the occupancy of adjacent nodes of the node where the current child node is located according to the occupancy of adjacent nodes of the node where the current child node is located;

[0021] Determine the occupancy context of the neighbor nodes of the current child node according to the context based on the adjacent nodes of the child node and the context based on the adjacent nodes of the node where the child node is located;

[0022] Determine the context of the current child node according to the occupancy context of the neighbor nodes of the current child node.

[0023] The point cloud geometry decoding method described above is characterized in that the step of determining the context of the current child node according to the combined occupancy number of adjacent nodes of the current child node includes:

[0024] Determine the occupancy context of the neighbor nodes of the current child node according to the combined occupancy number of adjacent nodes of the current child node;

[0025] Determine the occupancy context of the neighbor child nodes of the current child node according to the occupancy information of the encoded adjacent child nodes of the current child node;

[0026] Determine the context of the current child node according to the occupancy context of the neighbor nodes and the occupancy context of the neighbor child nodes of the current child node.

[0027] The point cloud geometry decoding method described above is characterized in that the step of determining the occupancy context of the neighbor child nodes of the current child node according to the occupancy information of the encoded adjacent child nodes of the current child node includes:

[0028] Obtain the information state M according to the occupancy information of the encoded adjacent child nodes of the current child node;

[0029] Convert the state M through a sliding window to obtain the state N, and use the state N as the occupancy context of the neighbor child nodes of the current child node.

[0030] A point cloud geometry decoding device is characterized by including a processor, a memory and a communication bus; a computer-readable program executable by the processor is stored on the memory;

[0031] The communication bus realizes the connection and communication between the processor and the memory;

[0032] When the processor executes the computer-readable program, it realizes the steps in the point cloud geometry decoding method described in any one of the above.

[0033] A point cloud geometric coding method, wherein the point cloud is defined in a tree structure, and each node in the tree structure includes a plurality of child nodes, characterized in that the method comprises the following steps:

[0034] Determine the context of the current child node according to the combined occupancy number of adjacent nodes of the current child node;

[0035] Perform entropy coding on the occupancy information of the current child node according to the context of the current child node to obtain a point cloud geometric code stream.

[0036] The point cloud geometric coding method as described above, characterized in that the step of determining the context of the current child node according to the combined occupancy number of adjacent nodes of the current child node includes:

[0037] Determine the context of the current child node according to the combined occupancy number of coplanar and collinear adjacent nodes of the current child node and according to the geometric structure of the adjacent nodes;

[0038] Alternatively, determine the context of the current child node according to the combined occupancy number of coplanar, collinear and concurrent adjacent nodes of the current child node and according to the geometric structure of the adjacent nodes.

[0039] A point cloud geometric coding device, characterized in that it includes a processor, a memory and a communication bus; a computer-readable program executable by the processor is stored on the memory;

[0040] The communication bus realizes the connection and communication between the processor and the memory;

[0041] When the processor executes the computer-readable program, the steps in the point cloud geometric coding method described in any one of the above are realized.

[0042] Compared with the prior art, the point cloud geometric coding method, decoding method, coding device and decoding device provided by the present invention have the following beneficial effects. The present invention determines the context of the current child node according to the combined occupancy number of adjacent nodes of the current child node. The combined occupancy number takes into account the occupancy numbers of adjacent nodes in multiple types of adjacent relationships of the current child node, and determines the context based on the combined occupancy number according to the geometric structure of the adjacent nodes. Compared with the prior art that only considers the occupancy number of a single adjacent relationship, the present invention takes into account the combined occupancy numbers of adjacent nodes in multiple types of adjacent relationships and the geometric structures of adjacent nodes under different numbers, and designs the context according to the combined occupancy number and geometric structure, making better use of the spatial distribution correlation of adjacent nodes and improving the point cloud geometric compression performance. Description of the Drawings

[0043] Figure 1 is a schematic flowchart of a point cloud geometric decoding method according to an embodiment of the present invention;

[0044] Figure 2 is another schematic flowchart of the point cloud geometry decoding method according to an embodiment of the present invention;

[0045] Figure 3 is another schematic flowchart of the point cloud geometry decoding method according to an embodiment of the present invention;

[0046] Figure 4 is a schematic diagram of the geometric structure when the combined occupancy number of adjacent nodes of the current child node according to an embodiment of the present invention is 1;

[0047] Figure 5 is another schematic diagram of the geometric structure when the combined occupancy number of adjacent nodes of the current child node according to an embodiment of the present invention is 1;

[0048] Figure 6 is a schematic diagram of the geometric structure when the combined occupancy number of adjacent nodes of the current child node according to an embodiment of the present invention is 2;

[0049] Figure 7 is a schematic diagram of the geometric structure when the combined occupancy number of adjacent nodes of the current child node according to an embodiment of the present invention is 3;

[0050] Figure 8 is a schematic diagram of the reduction of the geometric structure types of adjacent nodes of the current child node according to an embodiment of the present invention;

[0051] Figure 9 is a schematic diagram of the occupancy of adjacent nodes of the node where the current child node is located according to an embodiment of the present invention;

[0052] Figure 10 is a schematic diagram of the occupancy of the encoded adjacent child nodes of the current child node according to an embodiment of the present invention;

[0053] Figure 11 is another schematic diagram of the occupancy of the encoded adjacent child nodes of the current child node according to an embodiment of the present invention;

[0054] Figure 12 is a schematic flowchart of the point cloud geometry encoding method according to an embodiment of the present invention;

[0055] Figure 13 is a block diagram of the device structure according to an embodiment of the present invention. Detailed implementation manners

[0056] The present invention provides a point cloud geometry encoding method, a decoding method, an encoding device and a decoding device. To make the objectives, technical solutions and effects of the present invention clearer and more definite, the following further describes the present invention in detail with reference to the accompanying drawings and by way of examples. It should be understood that the specific examples described herein are only used to explain the present invention and are not used to limit the present invention.

[0057] Those skilled in the art of the present technology can understand that, unless specifically stated otherwise, the singular forms "a", "an", "the" and "said" used herein may also include the plural forms. It should be further understood that the term "comprising" used in the specification of the present invention means the presence of the described features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or their groups. It should be understood that when we say an element is "connected" or "coupled" to another element, it can be directly connected or coupled to other elements, or there may also be intermediate elements. In addition, the "connection" or "coupling" used herein may include wireless connection or wireless coupling. The phrase "and / or" used herein includes all or any unit and all combinations of one or more of the associated listed items.

[0058] Those skilled in the art of the present technology can understand that, unless otherwise defined, all terms used herein (including technical terms and scientific terms) have the same meaning as the general understanding of those of ordinary skill in the art to which the present invention pertains. It should also be understood that terms such as those defined in a general dictionary should be understood to have a meaning consistent with the meaning in the context of the prior art, and will not be interpreted with an idealized or overly formal meaning unless specifically defined as herein.

[0059] The following further describes the present invention through specific embodiments in conjunction with the accompanying drawings.

[0060] The specific usage scenario of the present invention is point cloud geometric encoding and geometric decoding based on an octree.

[0061] Please refer to Figure 1 , the present invention provides a method for point cloud geometric decoding. The point cloud is defined in a tree structure, and each node in the tree structure includes a plurality of child nodes, including the following steps:

[0062] D1: Determine the context of the current child node according to the combined occupancy number of adjacent nodes of the current child node;

[0063] D2: Entropy decode the point cloud geometric code stream according to the context of the current child node to obtain the occupancy information of the current child node.

[0064] In this embodiment, the steps D1 - D2 are executed at the decoding end to implement the geometric decoding process of the point cloud. According to the octree structure, each node in the tree structure includes 8 child nodes, and the occupancy codes corresponding to the 8 child nodes are (b7b6b5b4b3b2b1b0), and each child node corresponds to an occupancy information b k , k = 0, 1,... 7. If the child node does not contain any points, b k= 0; conversely, b k = 1.

[0065] Here, the context of the current child node is determined according to the combined occupancy number of the adjacent nodes of the current child node, and then the point cloud geometric code stream is entropy decoded according to the context of the current child node to obtain the occupancy information b of the current child node k .

[0066] Here, the adjacent nodes of the current child node refer to the nodes that are coplanar, collinear, or concurrent with the current child node.

[0067] In a possible implementation manner, the determining the context of the current child node according to the combined occupancy number of the adjacent nodes of the current child node includes:

[0068] D11: Determine the context of the current child node according to the combined occupancy number of the coplanar and collinear adjacent nodes of the current child node and according to the geometric structure of the adjacent nodes;

[0069] D12: Alternatively, determine the context of the current child node according to the combined occupancy number of the coplanar, collinear, and concurrent adjacent nodes of the current child node and according to the geometric structure of the adjacent nodes.

[0070] Here, the combined occupancy number refers to the combined occupancy number of at least two types of adjacent nodes in the adjacent relationships of coplanarity, collinearity, or concurrency of the current child node.

[0071] Figure 4 This is a schematic diagram of the geometric structure when the combined occupancy number of the adjacent nodes of the current child node in the embodiment of the present invention is 1. When the occupancy number combinations of the coplanar and collinear adjacent nodes of the current child node are combined and the combined occupancy number is 1, there are 2 cases of the geometric structure of the adjacent nodes, so 2 contexts are allocated. Figure 4 a is that there is 1 coplanar adjacent node for the adjacent node, Figure 4 b is that there is 1 collinear adjacent node for the adjacent node. Among them, a is the current child node, A is the node where the current child node is located, B is the coplanar adjacent node of the current child node, and C is the collinear adjacent node of the current child node.

[0072] Figure 5 This is another schematic diagram of the geometric structure when the combined occupancy number of the adjacent nodes of the current child node in the embodiment of the present invention is 1. When the occupancy number combinations of the coplanar, collinear, and concurrent adjacent nodes of the current child node are combined and the combined occupancy number is 1, there are 3 cases of the geometric structure of the adjacent nodes, so 3 contexts are allocated. Figure 5 a is that there is 1 coplanar adjacent node for the adjacent node, Figure 5 b is that there is 1 collinear adjacent node for the adjacent node, Figure 5c means that there is 1 common adjacent node among adjacent nodes. Among them, a is the current child node, A is the node where the current child node is located, B is the coplanar adjacent node of the current child node, C is the collinear adjacent node of the current child node, and D is the common adjacent node of the current child node.

[0073] In a possible implementation manner, determining the context of the current child node according to the combined occupancy number of adjacent nodes of the current child node and according to the geometric structure of the adjacent nodes includes:

[0074] Determining the context of the current child node according to the combined occupancy number of adjacent nodes of the current child node and according to all geometric structure types of adjacent nodes under the combined occupancy number;

[0075] Or, determining the context of the current child node according to the combined occupancy number of adjacent nodes of the current child node, determining all geometric structure types of adjacent nodes according to the combined occupancy number, and reducing the geometric structure types of adjacent nodes according to the number of coplanar adjacent nodes.

[0076] As an example, determining the context of the current child node according to the combined occupancy number of coplanar and collinear adjacent nodes of the current child node and according to all geometric structure types of adjacent nodes under the combined occupancy number. When the combined occupancy number is 0, there is only 1 geometric structure of adjacent nodes, that is, there are no coplanar and collinear adjacent nodes, and 1 context is allocated; when the combined occupancy number is 1, there are 2 geometric structures of adjacent nodes, that is, 1 coplanar and 1 collinear, as Figure 4 shown in a, and 2 contexts are allocated.

[0077] Figure 6 It is a schematic diagram of the geometric structure when the combined occupancy number of adjacent nodes of the current child node in the embodiment of the present invention is 2. When the combined occupancy number is 2, there are 4 geometric structures of adjacent nodes, that is, 1 type of 2 coplanar, as Figure 6 shown in a; 1 type of 2 collinear, as Figure 6 shown in b; 2 types of 1 coplanar and 1 collinear, as Figure 6 shown in c, 6d.

[0078] Figure 7 It is a schematic diagram of the geometric structure when the combined occupancy number of adjacent nodes of the current child node in the embodiment of the present invention is 3. When the combined occupancy number is 3, there are 6 geometric structures of adjacent nodes, that is, 1 type of 3 coplanar, as Figure 7 shown in a; 1 type of 3 collinear, as Figure 7 shown in d; 2 types of 2 coplanar and 1 collinear, as Figure 7 shown in b, 7e; 2 types of 1 coplanar and 2 collinear, as Figure 7c. As shown in 7f. Considering the symmetry between occupied nodes and unoccupied nodes, the combined occupancy numbers of 4, 5, 6 are the same as those of 2, 1, 0. The combined occupancy numbers, geometric structures, and contexts of the adjacent nodes of the current child node are shown in Table 1.

[0079] Table 1 Combined occupancy numbers, geometric structures, and contexts of the adjacent nodes of the current child node

[0080]

[0081] The occupancy number range of the adjacent nodes of the current child node is 0 - 6. The context determined according to the geometric structure of the adjacent nodes is represented by C1. The total number of C1 is 1 + 2 + 4 + 6 + 4 + 2 + 1 = 20, and the range of C1 is 0 - 19.

[0082] As an example, according to the combined occupancy number of the adjacent nodes of the current child node, all geometric structure types of the adjacent nodes are determined according to the combined occupancy number, and the geometric structure types of the adjacent nodes are reduced according to the number of coplanar adjacent nodes to determine the context of the current child node. Reducing the geometric structure types of the adjacent nodes according to the number of coplanar adjacent nodes can, on the one hand, reduce the corresponding context quantity, and on the other hand, merge the corresponding statistical probabilities. By reasonably merging the statistical probabilities, the entropy coding efficiency can be more efficient.

[0083] Figure 8 This is a schematic diagram of reducing the geometric structure types of the adjacent nodes of the current child node in the embodiment of the present invention. Consider the 3 geometric structures without coplanar adjacent nodes as 1 type for reduction, that is, consider the 3 geometric structures of the adjacent nodes of 1 - collinear, 2 - collinear, and 3 - collinear as 1 type for reduction, as shown in Figure 8 a; Consider the 3 geometric structures including 3 coplanar adjacent nodes as 1 type for reduction, that is, consider the 3 geometric structures of the adjacent nodes of 3 - coplanar 1 - collinear, 3 - coplanar 2 - collinear, and 3 - coplanar 3 - collinear as 1 type for reduction, as shown in Figure 8 b. In this way, after reduction, the original 20 geometric structures and corresponding contexts are reduced to 16, and the value range is 0 - 15.

[0084] Figure 2 This is another schematic flowchart of the point cloud geometry decoding method according to the embodiment of the present invention. In a possible implementation manner, determining the context of the current child node according to the combined occupancy number of the adjacent nodes of the current child node includes:

[0085] D1’: Determine the context based on the adjacent nodes of the child node according to the combined occupancy number of the adjacent nodes of the current child node;

[0086] D2’: Determine the context based on the adjacent nodes of the node where the current child node is located according to the occupancy of the adjacent nodes.

[0087] D3’: Determine the occupancy context of the neighbor nodes of the current child node according to the context based on the adjacent nodes of the child node and the context based on the adjacent nodes of the node where the child node is located.

[0088] D4’: Determine the context of the current child node according to the occupancy context of the neighbor nodes of the current child node.

[0089] Here, determine the context C1 based on the adjacent nodes of the current child node according to the combined occupancy number of the adjacent nodes of the current child node, determine the context C2 based on the adjacent nodes of the node where the current child node is located according to the occupancy of the adjacent nodes of the node where the current child node is located, and determine the occupancy context C=(C1)×2 + C2 of the neighbor nodes of the current child node according to C1 and C2, and determine the context of the current child node according to C.

[0090] Figure 9 It is a schematic diagram of the occupancy of the adjacent nodes of the node where the current child node is located in the embodiment of the present invention. Regarding the 6 coplanar adjacent nodes of the node where the current child node is located not being distributed in 3 axis directions as one geometric structure, assign a context 0, and regarding the 6 coplanar adjacent nodes of the node where the current child node is located being distributed in 3 axis directions as one geometric structure, assign a context 1.

[0091] Figure 3 It is another schematic flowchart of the point cloud geometry decoding method according to the embodiment of the present invention. In a possible implementation manner, determining the context of the current child node according to the combined occupancy number of the adjacent nodes of the current child node includes:

[0092] D1”: Determine the occupancy context of the neighbor nodes of the current child node according to the combined occupancy number of the adjacent nodes of the current child node.

[0093] D2”: Determine the occupancy context of the neighbor child nodes of the current child node according to the occupancy information of the encoded adjacent child nodes of the current child node.

[0094] D3”: Determine the context of the current child node according to the occupancy context of the neighbor nodes of the current child node and the occupancy context of the neighbor child nodes.

[0095] Here, according to the combined occupancy number of the adjacent nodes of the current child node, determine the neighbor node occupancy context C of the current child node; according to the occupancy information of the encoded adjacent child nodes of the current child node, determine the neighbor child node occupancy context N of the current child node; according to C and N, determine the context I of the current child node.

[0096] Figure 10 This is a schematic diagram of the occupancy of the encoded adjacent child nodes of the current child node in an embodiment of the present invention. Among them, the black sub-block is the current child node, and the gray sub-blocks are the coplanar, collinear, and concurrent adjacent child nodes of the current child node. One node contains 8 child nodes, and 8a-8h in the figure illustrate the coplanar, collinear, and concurrent adjacent child nodes at the positions of 8 child nodes.

[0097] Here, the encoded adjacent child nodes of the current child node include 7 adjacent child nodes in total, namely 3 coplanar, 3 collinear, and 1 concurrent. The corresponding occupancy information is represented by 7 bits, denoted as C3.

[0098] In a possible implementation manner, the determining the neighbor child node occupancy context of the current child node according to the occupancy information of the encoded adjacent child nodes of the current child node includes:

[0099] Obtain the information state M according to the occupancy information of the encoded adjacent child nodes of the current child node;

[0100] Convert the state M through a sliding window to obtain the state N, and use the state N as the neighbor child node occupancy context of the current child node.

[0101] Figure 11 This is another schematic diagram of the occupancy of the encoded adjacent child nodes of the current child node in an embodiment of the present invention. Among them, the black sub-block is the current child node, and the gray sub-blocks are the adjacent child nodes with an axial interval of 1 of the current child node. The corresponding occupancy information of 3 adjacent child nodes with an axial interval of 1 is represented by 3 bits, denoted as C4.

[0102] Here, the occupancy information M according to the encoded adjacent child nodes of the current child node includes C3 and C4, a total of 10 bits, with a value range of 0-1023 and a total of 1024 states. Record the encoded bit placeholder information in each state, and add the nearest K encoded bit placeholder information through a sliding window as the state N, and use the state N as the neighbor child node occupancy context of the current child node.

[0103]

[0104] Then the value range of the state N is 0 to K. If K is taken as 8, then N has a total of 9 state values.

[0105] Here, according to the context C occupied by the neighbor nodes of the current child node and the context N occupied by the neighbor child nodes, the context I of the current child node is determined as I = C × (K + 1) + N. Where K represents the size of the sliding window.

[0106] The present invention determines the context of the current child node according to the combined occupancy number of the adjacent nodes of the current child node. The combined occupancy number takes into account the occupancy numbers of the adjacent nodes of multiple types of adjacent relationships of the current child node, and determines the context based on the combined occupancy number according to the geometric structure of the adjacent nodes. Compared with the prior art that only considers the occupancy number of a single adjacent relationship, the present invention takes into account the combined occupancy numbers of the adjacent nodes of multiple types of adjacent relationships and the geometric structures of the adjacent nodes under different numbers. The context designed according to the combined occupancy number and the geometric structure makes better use of the spatial distribution correlation of the adjacent nodes and improves the point cloud geometry compression performance.

[0107] Please refer to Figure 12 , the present invention provides a point cloud geometry encoding method. The point cloud is defined in a tree structure, and each node in the tree structure includes multiple child nodes. The method includes the following steps:

[0108] S1: Determine the context of the current child node according to the combined occupancy number of the adjacent nodes of the current child node;

[0109] S2: Entropy encode the occupancy information of the current child node according to the context of the current child node to obtain a point cloud geometry code stream.

[0110] Here, step S1 is the same as step D1 and will not be described in detail.

[0111] Here, the occupancy information of the current child node is entropy encoded according to the context of the current child node to obtain a point cloud geometry code stream.

[0112] In a possible implementation manner, the determining the context of the current child node according to the combined occupancy number of the adjacent nodes of the current child node includes

[0113] S11: Determine the context of the current child node according to the combined occupancy number of the coplanar and collinear adjacent nodes of the current child node according to the geometric structure of the adjacent nodes;

[0114] S12: Alternatively, determine the context of the current child node according to the combined occupancy number of the coplanar, collinear, and concurrent adjacent nodes of the current child node according to the geometric structure of the adjacent nodes.

[0115] Here, step S11 is the same as step D11, and step S12 is the same as step D12, and will not be described in detail.

[0116] The present invention determines the context of the current child node according to the combined occupancy number of the adjacent nodes of the current child node. The combined occupancy number takes into account the occupancy numbers of the adjacent nodes of multiple types of adjacent relationships of the current child node, and determines the context based on the combined occupancy number according to the geometric structure of the adjacent nodes. Compared with the prior art that only considers the occupancy number of a single adjacent relationship, the present invention takes into account the combined occupancy numbers of the adjacent nodes of multiple types of adjacent relationships and the geometric structures of the adjacent nodes under different numbers. The context designed according to the combined occupancy number and the geometric structure makes better use of the spatial distribution correlation of the adjacent nodes and improves the point cloud geometric compression performance.

[0117] The following are the experimental results.

[0118] This experiment is based on the PCRM software version 9.0, and the experimental results of comparing the method of the present invention with the original sliding window method are tested.

[0119] Test conditions: lossy geometry, lossless geometry. The results are shown in Tables 2 and 3.

[0120] Table 2: Comparison results between the present invention and PCRM v9.0

[0121] Lossy Geometry Geometry Dataset 1 -1.0％ Dataset 2 -1.0％

[0122] Table 3: Comparison results between the present invention and PCRM v9.0

[0123] Lossless Geometry Geometry Dataset 1 99.9％ Dataset 2 96.8％

[0124] In Table 2, a 1.0% performance improvement can be seen under the condition of lossy geometry. In Table 3, a 3.2% performance improvement can be seen under the condition of lossless geometry.

[0125] Figure 13 It is the block diagram of the device according to the embodiment of the present invention. As Figure 13 shown, the device of the present invention includes a processor 20, a memory 22 and a communication bus 24; a computer-readable program executable by the processor is stored on the memory 22;

[0126] The communication bus 24 realizes the connection and communication between the processor 20 and the memory 22;

[0127] The communication bus 24 is connected to a communication interface 23, and when the processor 20 executes the computer-readable program, the steps in the point cloud geometric decoding method or the encoding method are realized.

[0128] The above embodiments are only specific implementation manners of the present invention, used to illustrate the technical solutions of the present invention, rather than limiting it. The protection scope of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: Any person skilled in the technical field of the present invention can still modify the technical solutions recorded in the foregoing embodiments, or can easily think of changes, or perform equivalent replacements on some of the technical features; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A point cloud geometric decoding method, wherein the point cloud is defined in a tree structure, and each node in the tree structure includes a plurality of child nodes, characterized in that, Including the following steps: Determine the context of the current child node according to the combined occupancy number of adjacent nodes of the current child node and the geometric structure of the adjacent nodes; The combined occupancy number refers to the combined occupancy number of at least two types of adjacent nodes in the coplanar, collinear, or concurrent adjacent relationships including the current child node; perform entropy decoding on the point cloud geometric code stream according to the context of the current child node to obtain the occupancy information of the current child node.

2. The point cloud geometry decoding method according to claim 1, wherein The determining the context of the current child node according to the combined occupancy number of adjacent nodes of the current child node and the geometric structure of the adjacent nodes includes: Determine the context of the current child node according to the combined occupancy number of coplanar and collinear adjacent nodes of the current child node and the geometric structure of the adjacent nodes; Alternatively, determine the context of the current child node according to the combined occupancy number of coplanar, collinear, and concurrent adjacent nodes of the current child node and the geometric structure of the adjacent nodes.

3. The point cloud geometry decoding method according to claim 2, wherein The determining the context of the current child node according to the combined occupancy number of adjacent nodes of the current child node and the geometric structure of the adjacent nodes includes: Determine the context of the current child node according to the combined occupancy number of adjacent nodes of the current child node and all geometric structure types of the adjacent nodes under the combined occupancy number; Alternatively, determine all geometric structure types of adjacent nodes according to the combined occupancy number of adjacent nodes of the current child node, reduce the geometric structure types of adjacent nodes according to the number of coplanar adjacent nodes, and determine the context of the current child node.

4. The point cloud geometry decoding method according to claim 1, characterized in that, The determining the context of the current child node according to the combined occupancy number of adjacent nodes of the current child node and the geometric structure of the adjacent nodes includes: Determine the context based on adjacent nodes of the child node according to the combined occupancy number of adjacent nodes of the current child node; Determine the context based on adjacent nodes of the node where the current child node is located according to the occupancy of adjacent nodes of the node where the current child node is located; Determine the occupancy context of neighbor nodes of the current child node according to the context based on adjacent nodes of the child node and the context based on adjacent nodes of the node where the current child node is located; Determine the context of the current child node according to the occupancy context of neighbor nodes of the current child node.

5. The point cloud geometry decoding method according to claim 1, characterized in that The determining the context of the current child node according to the combined occupancy number of adjacent nodes of the current child node and the geometric structure of the adjacent nodes includes: Determine the occupancy context of neighbor nodes of the current child node according to the combined occupancy number of adjacent nodes of the current child node; Determine the occupancy context of neighbor child nodes of the current child node according to the occupancy information of encoded adjacent child nodes of the current child node; Determine the context of the current child node according to the occupancy context of neighbor nodes and the occupancy context of neighbor child nodes of the current child node.

6. The point cloud geometry decoding method according to claim 5, characterized in that The determining the occupancy context of neighbor child nodes of the current child node according to the occupancy information of encoded adjacent child nodes of the current child node includes: Obtain the information state M according to the occupancy information of encoded adjacent child nodes of the current child node; The state M is transformed through a sliding window to obtain a state N, and the state N is used as the occupancy context of the neighbor child node of the current child node.

7. A point cloud geometry decoding device, characterized in that, It includes a processor, a memory, and a communication bus; a computer-readable program executable by the processor is stored on the memory; The communication bus realizes the connection and communication between the processor and the memory; When the processor executes the computer-readable program, it realizes the steps in the point cloud geometry decoding method described in any one of claims 1-6.

8. A point cloud geometric encoding method, wherein the point cloud is defined in a tree structure, and each node in the tree structure includes a plurality of child nodes, characterized in that, It includes the following steps: According to the combined occupancy number of adjacent nodes of the current child node, and in accordance with the geometric structure of the adjacent nodes, the context of the current child node is determined; The combined occupancy number refers to the combined occupancy number of at least two types of adjacent nodes in the coplanar, collinear, or concurrent adjacent relationships including the current child node; entropy coding is performed on the occupancy information of the current child node according to the context of the current child node to obtain a point cloud geometry bitstream.

9. The point cloud geometric coding method according to claim 8, wherein, The determining of the context of the current child node according to the combined occupancy number of adjacent nodes of the current child node and in accordance with the geometric structure of the adjacent nodes includes: According to the combined occupancy number of coplanar and collinear adjacent nodes of the current child node, and in accordance with the geometric structure of the adjacent nodes, the context of the current child node is determined; Or, according to the combined occupancy number of coplanar, collinear, and concurrent adjacent nodes of the current child node, and in accordance with the geometric structure of the adjacent nodes, the context of the current child node is determined.

10. A point cloud geometry encoding device, characterized in that, It includes a processor, a memory, and a communication bus; a computer-readable program executable by the processor is stored on the memory; The communication bus realizes the connection and communication between the processor and the memory; When the processor executes the computer-readable program, it realizes the steps in the point cloud geometry encoding method described in any one of claims 8-9.

Images