WO2025138196A1 - Encoding method, decoding method, point cloud encoder, point cloud decoder, bitstream, and storage medium - Google Patents

Abstract

Provided are an encoding method, a decoding method, a point cloud encoder, a point cloud decoder, a bitstream, and a storage medium. The decoding method comprises: decoding first identification information on the basis of a first probability model, wherein the first identification information is used for indicating whether the current node contains duplicated points; and if the first identification information indicates that the current node contains duplicated points, determining the number of duplicated points contained in the current node. The first probability model satisfies at least one of the following: an initial value of the first probability model being not equal to a first value, which corresponds to the probability of an MPS being equal to the probability of an LPS; and the first probability model being different from a second probability model, the second probability model corresponding to second identification information. The first identification information is used for indicating whether a node using DCM contains duplicated points, and the second identification information is used for indicating whether a leaf node of a multi-way tree contains duplicated points. Alternatively, the first identification information is used for indicating whether the leaf node of the multi-way tree contains duplicated points, and the second identification information is used for indicating whether the node using the DCM contains duplicated points.

Description

Coding and decoding method, point cloud codec, code stream and storage medium Technical Field

The present application relates to the field of video coding and decoding technology, and in particular to a coding and decoding method, a point cloud codec, a bit stream and a storage medium.

Background Art

In the geometry-based point cloud compression (G-PCC) coding and decoding framework, the geometric information and attribute information of the point cloud are encoded and decoded separately. In the geometric coding process, the design of the coding and decoding scheme for repeated points is unreasonable, which reduces the coding efficiency.

Summary of the invention

The present application provides a coding and decoding method, a point cloud codec, a bit stream and a storage medium. The following is an introduction to various aspects of the present application.

In a first aspect, a decoding method is provided, which is applied to a point cloud decoder, comprising: decoding first identification information according to a first probability model, the first identification information being used to indicate whether a current node contains duplicate points; if the first identification information indicates that the current node contains duplicate points, determining the number of duplicate points contained in the current node; wherein the first probability model satisfies at least one of the following: an initial value of the first probability model is not equal to a first value, the first value corresponding to: the probabilities of a maximum probability symbol (most probability symbol, MPS) and a minimum probability symbol (low probability symbol, LPS) are equal; the first probability model is different from a second probability model, the second probability model is a probability model corresponding to the second identification information; wherein the first identification information is used to indicate whether a node using a direct coding mode contains duplicate points, and the second identification information is used to indicate whether a leaf node of a multi-branch tree contains duplicate points; or, the first identification information is used to indicate whether a leaf node of a multi-branch tree contains duplicate points, and the second identification information is used to indicate whether a node using a direct coding mode contains duplicate points.

In a second aspect, a coding method is provided, which is applied to a point cloud encoder, comprising: encoding first identification information according to a first probability model, the first identification information being used to indicate whether a current node contains duplicate points; if the first identification information indicates that the current node contains duplicate points, encoding the number of duplicate points contained in the current node; wherein the first probability model satisfies at least one of the following: an initial value of the first probability model is not equal to a first value, and the first value corresponds to: the probabilities of MPS and LPS are equal; the first probability model is different from the second probability model, and the second probability model is a probability model corresponding to the second identification information; wherein the first identification information is used to indicate whether a node using a direct coding mode contains duplicate points, and the second identification information is used to indicate whether a leaf node of a multi-branch tree contains duplicate points; or, the first identification information is used to indicate whether a leaf node of a multi-branch tree contains duplicate points, and the second identification information is used to indicate whether a node using a direct coding mode contains duplicate points.

According to a third aspect, a point cloud decoder is provided, comprising: a decoding unit, configured to decode first identification information according to a first probability model, the first identification information being used to indicate whether a current node contains duplicate points; a determination unit, configured to determine the number of duplicate points contained in the current node if the first identification information indicates that the current node contains duplicate points; wherein the first probability model satisfies at least one of the following: an initial value of the first probability model is not equal to a first value, and the first value corresponds to: the probabilities of MPS and LPS are equal; the first probability model is different from the second probability model, and the second probability model is a probability model corresponding to the second identification information; wherein the first identification information is used to indicate whether a node using a direct coding mode contains duplicate points, and the second identification information is used to indicate whether a leaf node of a multi-branch tree contains duplicate points; or, the first identification information is used to indicate whether a leaf node of a multi-branch tree contains duplicate points, and the second identification information is used to indicate whether a node using a direct coding mode contains duplicate points.

In a fourth aspect, a point cloud decoder is provided, comprising: a memory for storing a computer program; and a processor for executing the method of the first aspect when running the computer program.

In a fifth aspect, a point cloud encoder is provided, comprising: a first encoding unit, configured to encode first identification information according to a first probability model, the first identification information being used to indicate whether a current node contains duplicate points; a second encoding unit, configured to encode the number of duplicate points contained in the current node if the first identification information indicates that the current node contains duplicate points; wherein the first probability model satisfies at least one of the following: an initial value of the first probability model is not equal to a first value, and the first value corresponds to: the probabilities of MPS and LPS are equal; the first probability model is different from the second probability model, and the second probability model is a probability model corresponding to the second identification information; wherein the first identification information is used to indicate whether a node using a direct encoding mode contains duplicate points, and the second identification information is used to indicate whether a leaf node of a multi-branch tree contains duplicate points; or, the first identification information is used to indicate whether a leaf node of a multi-branch tree contains duplicate points, and the second identification information is used to indicate whether a node using a direct encoding mode contains duplicate points.

In a sixth aspect, a point cloud encoder is provided, comprising: a memory for storing a computer program; and a processor for executing the method of the second aspect when running the computer program.

In a seventh aspect, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed, the method of any one of the first to second aspects is implemented.

In an eighth aspect, a computer program product is provided, comprising a computer program, which implements the method of any one of the first to second aspects when the computer program is executed.

In a ninth aspect, a non-volatile computer-readable storage medium for storing a bit stream is provided, wherein the bit stream is generated by an encoding method using a point cloud encoder, or the bit stream is decoded by a decoding method using a point cloud decoder, wherein the decoding method is the method described in the first aspect, and the encoding method is the method described in the second aspect.

According to a tenth aspect, a code stream is provided, comprising a code stream generated according to the method described in the second aspect.

The embodiment of the present application adjusts the probability model of identification information related to repeated points. The adjusted probability model is more in line with the actual situation and helps to improve the encoding and decoding efficiency of the point cloud.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1A is a schematic diagram of a three-dimensional point cloud image.

FIG. 1B is a partially enlarged view of a three-dimensional point cloud image.

FIG. 2A is a schematic diagram of six viewing angles of a point cloud image.

FIG. 2B is a schematic diagram of a data storage format corresponding to a point cloud image.

FIG3 is a schematic diagram of a network architecture for point cloud encoding and decoding.

FIG. 4A is a schematic diagram of a composition framework of a G-PCC encoder.

FIG. 4B is a schematic diagram of a composition framework of a G-PCC decoder.

Figure 5 is an example diagram of the direct coding model (DCM).

FIG6 is an example diagram of the inference process of the infer direct coding model (IDCM).

FIG. 7 is a schematic diagram of the updating process of the probability model.

FIG8 is a schematic diagram of the renormalization process.

FIG. 9 is a flowchart of a decoding method provided by an embodiment of the present application.

FIG10 is a flowchart of an encoding method provided by an embodiment of the present application.

FIG. 11 is a flowchart of a decoding method provided in another embodiment of the present application.

FIG. 12 is a flowchart of an encoding method provided in another embodiment of the present application.

FIG13 is a schematic diagram of the structure of a point cloud decoder provided in one embodiment of the present application.

FIG. 14 is a schematic diagram of the structure of a point cloud decoder provided in another embodiment of the present application.

FIG. 15 is a schematic diagram of the structure of a point cloud encoder provided by an embodiment of the present application.

FIG16 is a schematic diagram of the structure of a point cloud encoder provided in another embodiment of the present application.

DETAILED DESCRIPTION

In order to enable a more detailed understanding of the features and technical contents of the embodiments of the present application, the implementation of the embodiments of the present application is described in detail below in conjunction with the accompanying drawings. The attached drawings are for reference only and are not used to limit the embodiments of the present application.

Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

In the following description, reference is made to “some embodiments” or “some implementations”, which describe a subset of all possible embodiments, but it can be understood that “some embodiments” or “some implementations” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

It should also be pointed out that the terms "first\second\third" involved in the embodiments of the present application are only used to distinguish similar objects and do not represent a specific ordering of the objects. It can be understood that "first\second\third" can be interchanged in a specific order or sequence where permitted, so that the embodiments of the present application described here can be implemented in an order other than that illustrated or described here.

The embodiments of the present application can be applied to point cloud encoding and decoding. For ease of understanding, the point cloud and its encoding and decoding method are first introduced below.

Point cloud is a three-dimensional representation of the surface of an object. Point cloud of the surface of an object can be collected through acquisition equipment such as photoelectric radar, lidar, laser scanner, and multi-view camera.

A point cloud is a set of irregularly distributed discrete points in space that express the spatial structure and surface properties of a three-dimensional object or scene. FIG1A shows a three-dimensional point cloud image. FIG1B shows a partial magnified view of the three-dimensional point cloud image. It can be seen from FIG1A and FIG1B that the point cloud surface is composed of densely distributed points.

Two-dimensional images have information expressed at each pixel point, and the distribution is regular, so there is no need to record its position information separately. However, the distribution of points in the point cloud in three-dimensional space is random and irregular, so it is necessary to record the position of each point in space in order to fully express a point cloud. Similar to two-dimensional images, each position in the acquisition process has corresponding attribute information, usually RGB color values, and the color value reflects the color of the object. For point clouds, in addition to color information, the attribute information corresponding to each point is also commonly the reflectance value, which reflects the surface material of the object. Therefore, point cloud data usually includes the location information of the point And the attribute information of the point. The location information of the point can also be called the geometric information of the point. For example, the geometric information of the point can be the three-dimensional coordinate information (x, y, z) of the point. The attribute information of the point can include color information and/or reflectivity, etc. For example, the reflectivity can be one-dimensional reflectivity information (r); the color information can be information on any color space, or the color information can be three-dimensional color information, such as RGB information. Here, R represents red (red, R), G represents green (green, G), and B represents blue (blue, B). For another example, the color information can be brightness and chromaticity (YCbCr, YUV) information. Among them, Y represents brightness (luma), Cb (U) represents blue color difference, and Cr (V) represents red color difference.

For a point cloud obtained according to the principle of laser measurement, the points in the point cloud may include the three-dimensional coordinate information of the points and the reflectivity value of the points. For another example, for a point cloud obtained according to the principle of photogrammetry, the points in the point cloud may include the three-dimensional coordinate information of the points and the three-dimensional color information of the points. For another example, a point cloud obtained by combining the principles of laser measurement and photogrammetry may include the three-dimensional coordinate information of the points, the reflectivity value of the points and the three-dimensional color information of the points.

Figures 2A and 2B show a point cloud image and its corresponding data storage format. Figure 2A provides six viewing angles of the point cloud image, and Figure 2B consists of a file header information part and a data part. The header information includes the data format, data representation type, the total number of point cloud points, and the content represented by the point cloud. For example, the point cloud is in the ".ply" format, represented by ASCII code, with a total number of 207242 points, and each point has three-dimensional coordinate information (x, y, z) and three-dimensional color information (r, g, b).

Point clouds can be divided into the following types according to the acquisition method:

Static point cloud: the object is stationary, and the device that obtains the point cloud is also stationary;

Dynamic point cloud: The object is moving, but the device that obtains the point cloud is stationary;

Dynamic point cloud acquisition: The device used to acquire the point cloud is in motion.

For example, point clouds can be divided into two categories according to their usage:

Category 1: Machine perception point cloud, which can be used in autonomous navigation systems, real-time inspection systems, geographic information systems, visual sorting robots, disaster relief robots, etc.

Category 2: Point cloud perceived by the human eye, which can be used in point cloud application scenarios such as digital cultural heritage, free viewpoint broadcasting, 3D immersive communication, and 3D immersive interaction.

Point clouds can flexibly and conveniently express the spatial structure and surface properties of three-dimensional objects or scenes. Point clouds are obtained by directly sampling real objects, so they can provide a strong sense of reality while ensuring accuracy. Therefore, they are widely used, including virtual reality games, computer-aided design, geographic information systems, automatic navigation systems, digital cultural heritage, free viewpoint broadcasting, three-dimensional immersive remote presentation, and three-dimensional reconstruction of biological tissues and organs.

Point clouds can be collected mainly through the following methods: computer generation, 3D laser scanning, 3D photogrammetry, etc. Computers can generate point clouds of virtual three-dimensional objects and scenes; 3D laser scanning can obtain point clouds of static real-world three-dimensional objects or scenes, and can obtain millions of point clouds per second; 3D photogrammetry can obtain point clouds of dynamic real-world three-dimensional objects or scenes, and can obtain tens of millions of point clouds per second. These technologies reduce the cost and time cycle of point cloud data acquisition and improve the accuracy of data. The change in the way point cloud data is acquired makes it possible to acquire a large amount of point cloud data. With the growth of application demand, the processing of massive 3D point cloud data encounters bottlenecks in storage space and transmission bandwidth.

For example, taking a point cloud video with a frame rate of 30 frames per second (fps) as an example, the number of points in each point cloud frame is 700,000, and each point has coordinate information xyz (float) and color information RGB (uchar). The data volume of a 10s point cloud video is about 0.7 million × (4Byte × 3 + 1Byte × 3) × 30fps × 10s = 3.15GB, where 1Byte is 10bit; and a 1280 × 720 two-dimensional video with a YUV sampling format of 4:2:0 and a frame rate of 24fps, the data volume of 10s is about 1280 × 720 × 12bit × 24fps × 10s ≈ 0.33GB, and the data volume of a 10s two-view three-dimensional video is about 0.33 × 2 = 0.66GB. It can be seen that the data volume of a point cloud video far exceeds that of a two-dimensional video and a three-dimensional video of the same length. Therefore, in order to better realize data management, save server storage space, and reduce the transmission traffic and transmission time between the server and the client, point cloud compression has become a key issue in promoting the development of the point cloud industry. In other words, since the point cloud is a collection of massive points, storing the point cloud will not only consume a lot of memory, but also be unfavorable for transmission. There is also no large bandwidth to support the point cloud to be transmitted directly at the network layer without compression. Therefore, the point cloud needs to be compressed.

At present, the point cloud coding framework that can compress point clouds can be the geometry-based point cloud compression (G-PCC) codec framework or the video-based point cloud compression (V-PCC) codec framework provided by the moving picture experts group (MPEG), or the AVS-PCC codec framework provided by the audio video coding standard (AVS). The G-PCC codec framework can be used to compress the first type of static point cloud and the third type of dynamically acquired point cloud, which can be based on the point cloud compression test platform (test model compression 13, TMC13), and the V-PCC codec framework can be used to compress the second type of dynamic point cloud, which can be based on the point cloud compression test platform (test model compression 2, TMC2). Therefore, the G-PCC codec framework is also called the point cloud codec TMC13, and the V-PCC codec framework is also called the point cloud codec TMC2.

FIG3 shows a schematic diagram of a network architecture for point cloud coding and decoding. As shown in FIG3 , the network architecture includes one or more electronic devices 13 to 1N and a communication network 01, wherein the electronic devices 13 to 1N can perform video interaction through the communication network 01. In the implementation process, various types of devices with point cloud encoding and decoding functions can be used, for example, the electronic device can include a mobile phone, a tablet computer, a personal computer, a personal digital assistant, a navigator, a digital phone, a video phone, a television, a sensor device, a server, etc., and the embodiments of the present application are not limited thereto. Among them, the decoder or encoder in the embodiments of the present application can be the above-mentioned electronic device.

Among them, the electronic device in the embodiment of the present application has a point cloud encoding and decoding function, generally including a point cloud encoder (ie, encoder) and a point cloud decoder (ie, decoder).

The following uses the G-PCC encoding and decoding framework as an example to illustrate the point cloud encoding and decoding method.

In the point cloud G-PCC encoding and decoding framework, the point cloud data to be encoded is first divided into multiple slices by slice division. In each slice, the geometric information and attribute information of the point cloud are encoded separately.

FIG4A shows a schematic diagram of the encoding framework of a G-PCC encoder. As shown in FIG4A , in the geometric encoding process, the geometric information is transformed so that all the point clouds are contained in a bounding box, and then quantized. Quantization mainly plays a role in scaling. Due to the quantization rounding, the geometric information of a part of the point cloud is the same, so whether to remove duplicate points is determined based on parameters. The process of quantization and removal of duplicate points is also called voxelization. Then, the bounding box is divided into octrees or a prediction tree is constructed. In this process, arithmetic coding is performed on the points in the leaf nodes of the division to generate a binary geometric bit stream; or, arithmetic coding is performed on the intersections (vertex) generated by the division (surface fitting is performed based on the intersections) to generate a binary geometric bit stream. In the attribute encoding process, after the geometric encoding is completed and the geometric information is reconstructed, color conversion is required first to convert the color information (i.e., attribute information) from the RGB color space to the YUV color space. Then, the point cloud is recolored using the reconstructed geometric information so that the unencoded attribute information corresponds to the reconstructed geometric information. Attribute encoding is mainly performed on color information. In the process of color information encoding, there are two main transformation methods: one is the distance-based lifting transformation that relies on the level of detail (LOD) division, and the other is the direct regional adaptive hierarchical transformation (RAHT). Both methods convert color information from the spatial domain to the frequency domain, obtain high-frequency coefficients and low-frequency coefficients through transformation, and finally quantize the coefficients. Then, the quantized coefficients are arithmetically encoded to generate a binary attribute bit stream.

FIG4B shows a schematic diagram of a decoding framework of a G-PCC decoder. As shown in FIG4B , for the acquired binary bit stream, the geometric bit stream and the attribute bit stream in the binary bit stream are first decoded independently. When decoding the geometric bit stream, the geometric information of the point cloud is obtained through arithmetic decoding-reconstruction of the octree/reconstruction of the prediction tree-reconstruction of the geometry-coordinate inverse conversion; when decoding the attribute bit stream, the attribute information of the point cloud is obtained through arithmetic decoding-inverse quantization-LOD partitioning/RAHT-color inverse conversion, and the point cloud data to be encoded (i.e., the output point cloud) is restored based on the geometric information and the attribute information.

According to the introduction of FIG. 4A and FIG. 4B , the current geometric coding and decoding of G-PCC can be divided into: octree-based geometric coding and decoding, triangle soup (trisoup)-based geometric coding and decoding, and prediction tree-based geometric coding and decoding. The specific implementation methods of the three geometric coding and decoding methods are introduced in more detail below.

Octree-based geometry encoding and decoding

At the encoding end, the geometric information of the point cloud is first transformed into coordinates so that all the point clouds are contained in a bounding box defined by two extreme points (0, 0, 0) and ( ^2d , ^2d , ^2d ). Then, the points in the bounding box are voxelized, that is, quantized, rounded, and duplicate points are removed (whether to remove duplicate points can be determined by parameters). Next, the non-empty nodes in the bounding box (or non-empty sub-cubes, that is, sub-cubes containing point clouds) are continuously divided into octrees in a breadth-first traversal order. At the same octree depth, a node will be divided into 8 child nodes until the leaf node obtained by the division is a 1x1x1 unit cube. Whether there is point occupancy in the node (1 for occupation and 0 for non-occupancy) can be represented by an 8-bit binary code. This 8-bit binary code is called an occupancy code. The placeholder code of each node is encoded to generate a binary code stream.

At the decoding end, the placeholder code of each node is continuously parsed in a breadth-first traversal order, and the nodes are continuously divided in turn until a 1x1x1 unit cube is obtained. Then, the number of points contained in each leaf node is parsed to restore the reconstructed geometric information of the point cloud.

Geometry encoding and decoding based on trisoup

At the encoding end, the division is first performed based on the octree. Different from the geometric information encoding based on the octree structure, the geometric encoding based on trisoup does not need to divide the point cloud into leaf nodes with a side length of 1x1x1 step by step, but divides it into leaf nodes with a specified side length. After reaching the leaf node with a specified side length, the surface information composed of the voxels in the node can be represented by a series of triangle meshes. In G-PCC, the parameter trisoup node size is used to represent the size of the block where the triangle patch is located. When the trisoup node size is greater than 0, the voxel set in the node is represented by a geometric patch. The up to twelve intersections generated by the geometric patch and the twelve edges of the block are called vertices. The vertex coordinates of each block are encoded in turn to generate a binary code stream.

At the decoding end, in order to decode the geometric information of the point cloud from the node triangle patch, it is necessary to check whether each voxel in the node cube intersects with the triangle patch. This technology is called triangle rasterization. For example, six unit vectors (0,0,1), (0,0,1), (0,0,1), (0,0,1), (0,0,1), (0,0,1) can be used for intersection check to check whether each unit vector intersects with the triangle patch. If they intersect, the intersection point is calculated and the decoded cube is output. The number of points generated in the decoder is determined by the grid distance d.

Geometric coding and decoding based on prediction tree

At the encoding end, the input point cloud is first sorted. The sorting methods currently used include unordered, Morton order, azimuth order, and radial distance order. At the encoding end, two different methods can be used to establish the prediction tree structure. One method is to build a prediction tree based on KD-Tree. This prediction tree construction method can also be called high-latency slow mode. Another method is to use the lidar calibration information to divide each point into different Lasers and establish a prediction structure according to different Lasers. This prediction tree construction method can also be called low-latency fast mode. Next, based on the structure of the prediction tree, traverse each node in the prediction tree, predict the geometric position information of the node by selecting different prediction modes to obtain the prediction residual, and use the quantization parameter to quantize the geometric prediction residual. Finally, through continuous iteration, the prediction residual of the prediction tree node position information, the prediction tree structure, and the quantization parameters are encoded to generate a binary code stream.

At the decoding end, the decoder continuously parses the bitstream to reconstruct the prediction tree structure. Secondly, the decoder obtains the geometric position prediction residual information and quantization parameters of each prediction node through parsing, and dequantizes the prediction residual to recover the reconstructed geometric position information of each node, and finally completes the geometric reconstruction at the decoding end.

The above article introduces the basic concept of point cloud and the basic method of point cloud encoding and decoding. The embodiment of the present application mainly improves the processing method of duplicate points in point cloud encoding and decoding. The following article introduces the technology related to duplicate points in the point cloud encoding and decoding process in detail.

Quantization and duplicate point removal (voxelization) in G-PCC

In G-PCC, the geometric preprocessing stage includes a key step, namely voxelization. In the voxelization process, the position of the point is represented as a non-negative integer before compression. In order to obtain these integer values, the position of the point is rounded. Let the initial position of the point be The quantification method of this point is as follows:

Round(.) means rounding the value in the brackets to the nearest integer.

After quantization, there may be multiple points with the same position, which are called duplicate points. The removal of duplicate points is optional and is controlled by the syntax element geom_unique_points_flag of the geometry parameter set (GPS) layer. When unique_points_flag is set to 0, the encoder will encode the number of duplicate points. Similarly, the decoder will decode the number of duplicate points only if the decoded unique_points_flag is equal to 0.

Therefore, the voxelization process in geometric preprocessing can be regarded as a process of replacing all points within a bounding box of [i-0.5,i+0.5)×[j-0.5,j+0.5)×[k-0.5,k+0.5) (where i, j, k are integer values between 0 and ^2d -1) with the center point of the bounding box.

Direct coding model (DCM)

As shown in Figure 5, in G-PCC, entropy coding of point coordinates in a node or subnode is called DCM. DCM is mainly used to process isolated nodes. The octree partitioning process uses neighbor context information for entropy coding. Unlike the octree partitioning process, DCM directly encodes the three coordinate components of the point.

For the node to be encoded, whether to enable DCM needs to be inferred through the parent node and its neighbor information. This process is called infer direct coding model (IDCM). As shown in Figure 6, IDCM will first determine whether the current node is eligible for DCM encoding. If the current node is eligible for DCM encoding, the current node can be DCM encoded (of course, other judgment conditions can also be introduced, such as judging whether the number of points in the current node is less than the threshold (nb_points≤th), etc.). If the current node is not eligible for DCM encoding, the current node can be octree encoded.

Whether the current node is eligible for DCM coding may be determined based on the parent node of the current node itself or the neighbor information of the parent node, which may be referred to as inference.

The inferences of DCM coding eligibility can be divided into the following two categories:

1) Parent-based eligibility: There is only one occupied child node at the parent node level (i.e. the current node), and the grandparent node (i.e. the parent node's parent node) has at most two occupied child nodes (i.e. the parent node and possibly one other node).

2) 6N qualification: There is only one occupied child node (i.e. the current node) at the parent node level, and no occupied neighbor nodes (the neighbor nodes mentioned here refer to the 6 neighbor nodes above, below, left, right, front and back).

In order to adapt to different compression efficiency and/or encoding time, IDCM has three possible inference methods. Inference method 1 means: if the current node meets the parent-based qualifications and 6N qualifications, DCM encoding is turned on. This inference method has the best compression effect, but the compression time is longer. Inference method 2 means: if the current node meets the 6N qualifications, DCM encoding is turned on. This inference method 2 relaxes the judgment conditions of DCM, and more nodes will enable DCM. Inference method 3 means: if the current node meets the parent-based qualifications, DCM encoding is turned on. This inference method uses more relaxed judgment conditions, but the compression efficiency is lower.

If the current node meets the DCM encoding qualification, a binary flag is first encoded to indicate whether the current node uses DCM. If the flag is 1, it means that the current node uses DCM; if the flag is 0, it means that the current node does not use DCM.

If the DCM mode is adopted, the following steps can be used to encode the current node.

First: Encode the binary syntax element direct_point_cnt_eq2. direct_point_cnt_eq2 indicates whether the current node contains two different points (i.e., the geometric information of the two points is different). If direct_point_cnt_eq2 is 1, it means that the current node contains two different points. points; if direct_point_cnt_eq2 is 0, it means that the current node only contains points with the same geometric information (it can be only 1 point or multiple points with the same geometric information).

Second: If direct_point_cnt_eq2 is 0, encode the number of duplicate points. First, encode the binary syntax element DupPointsCntGt0 to identify whether the current node contains duplicate points. If DupPointsCntGt0 is 1, it means that the current node contains duplicate points; if DupPointsCntGt0 is 0, it means that the current node does not contain duplicate points, and the number of duplicate points is 0. If DupPointsCntGt0 is 1, further encode the binary syntax element DupPointsCntGt1. DupPointsCntGt1 is used to identify whether the current node contains 1 duplicate point (that is, the geometric information of the two points is the same). If DupPointsCntGt1 is 1, it means that the current node contains more than 1 duplicate point; if DupPointsCntGt1 is 0, it means that the current node contains 1 duplicate point. If DupPointsCntGt1 is 1, further encode the syntax element DupPointsCntEg1. DupPointsCntEg1+2 is the number of duplicate points.

Encoding and decoding of duplicate points of multi-branch tree leaf nodes

In the multi-tree (such as octree) encoding of G-PCC, the multi-tree leaf node may contain duplicate points. The encoding method of the number of duplicate points of the multi-tree leaf node includes the following steps. First, encode the binary syntax element DupPointsCntGt0L. DupPointsCntGt0L can be used to indicate whether the leaf node contains duplicate points. If DupPointsCntGt0L is 1, it means that the leaf node contains duplicate points; if DupPointsCntGt0L is 0, it means that the leaf node does not contain duplicate points. Then, when DupPointsCntGt0L is 1, the number of duplicate points in the leaf node is encoded using Exponential Columbus.

The decoding method of the number of repeated points of a multi-fork tree leaf node includes the following steps. First, decode the binary syntax element DupPointsCntGt0L. If DupPointsCntGt0L is 1, it means that the leaf node contains repeated points; if DupPointsCntGt0L is 0, it means that the leaf node does not contain repeated points. When DupPointsCntGt0L is 1, continue to use the exponential Columbus decoding to decode the number of repeated points in the leaf node.

Entropy coding probability model and initialization

For a probability model, first determine the interval starting point (Low) and interval width (Range) of the corresponding arithmetic encoder. Then, the interval size of MPS and LPS can be continuously updated according to the input binary (bin) value, and finally the value of Low is used as the output of the encoding.

The updating process of the probability model is described in detail below in conjunction with FIG8 .

First, initialize Low and Range (the initial interval width of Range is 510, represented by 9 bits). If the initial value of the probability model is the first value (the first value corresponds to the case where the probabilities of MPS and LPS are equal, that is, if the initial value of the probability model is the first value, the probabilities of MPS and LPS are equal, both 0.5), then the width of the initial interval R _MPS of MPS is consistent with the initial interval width R _LPS of LPS, each occupying half of the initial interval width of Range.

Secondly, determine whether the Bin value to be encoded is equal to LPS or MPS, so as to update Low and Range. Referring to FIG7 , if Bin=LPS, then Low=Low+R _MPS , Range=R _LPS ; if Bin=MPS, then Low remains unchanged and Range is equal to R _MPS .

Then, the probability model is updated using Bin value, Low, and Range.

As Range is updated, the value of Range may be less than half of the initial interval length, that is, less than 256. At this time, a renormalization process is required. Figure 8 shows the process of the renormalization process. During the renormalization process, Low and Range are left shifted until the value of Range is greater than half of the initial interval width. The left shift of Range is to expand the interval width of the arithmetic encoder, and the left shift of Low is to output bits. The number of bits of the left shift of the two is the same.

The previous article introduced point cloud encoding and decoding and the processing of duplicate points in point cloud encoding and decoding in detail. After careful study, it was found that some algorithm designs of point cloud encoding and decoding are still unreasonable, which reduces the encoding and decoding efficiency. This is analyzed in detail below.

As mentioned above, the syntax element DupPointsCntGt0 in the DCM encoding is used to identify whether the node using DCM contains repeated points, and the syntax element DupPointsCntGt0L in the multi-fork tree leaf node is used to identify whether the multi-fork tree leaf node contains repeated points. In the related art, DupPointsCntGt0 and DupPointsCntGt0L use the same probability model for entropy coding. However, the actual probability that the node using DCM encoding and decoding contains repeated points is different from the actual probability that the multi-fork tree leaf node contains repeated points. The two use the same probability model for entropy coding, which is inconsistent with the actual situation, and therefore will cause a loss in coding efficiency.

In addition, the probability that the syntax element DupPointsCntGt0 in DCM coding and the syntax element DupPointsCntGt0L in the multi-fork tree leaf node take the value of 0 is very high. That is, in most cases, the nodes and multi-fork tree leaf nodes using DCM encoding and decoding do not include repeated points. The relevant technology does not make full use of this prior information, and sets the initial value of the probability model corresponding to DupPointsCntGt0 (sharing the same probability model with DupPointsCntGt0L) to the first value (the first value corresponds to the case where the probabilities of MPS and LPS are equal, that is, if the initial value of the probability model is the first value, the probabilities of MPS and LPS are equal, both 0.5). If the initial value of the probability model is set to the first value, the initial interval widths corresponding to MPS and LPS are consistent. At this time, the encoder needs to continuously output bits (output bin value 0) to update the interval width corresponding to MPS, so that the interval width of MPS reaches a certain degree. Only when the MPS interval width reaches the interval width corresponding to MPS close to 1, bits begin to be saved. The output of these bits is actually caused by unreasonable initial value setting. If the initial value can be set appropriately, the process of "continuously outputting bits to make the MPS interval reach a certain level" mentioned above can be avoided, thereby improving the coding efficiency.

In response to the above problems, an embodiment of the present application proposes a coding method, including: encoding first identification information according to a first probability model, the first identification information being used to indicate whether the current node contains duplicate points; if the first identification information indicates that the current node contains duplicate points, encoding the number of duplicate points contained in the current node; wherein the first probability model satisfies at least one of the following: an initial value of the first probability model is not equal to a first value, and the first value corresponds to: the probabilities of MPS and LPS are equal; the first probability model is different from the second probability model, and the second probability model is a probability model corresponding to the second identification information; wherein the first identification information is used to indicate whether a node using DCM contains duplicate points, and the second identification information is used to indicate whether a leaf node of a multi-branch tree contains duplicate points; or, the first identification information is used to indicate whether a leaf node of a multi-branch tree contains duplicate points, and the second identification information is used to indicate whether a node using a direct coding mode contains duplicate points.

An embodiment of the present application also proposes a decoding method, comprising: decoding first identification information according to a first probability model, the first identification information being used to indicate whether the current node contains duplicate points; if the first identification information indicates that the current node contains duplicate points, determining the number of duplicate points contained in the current node; wherein the first probability model satisfies at least one of the following: an initial value of the first probability model is not equal to a first value, and the first value corresponds to: the probabilities of MPS and LPS are equal; the first probability model is different from the second probability model, and the second probability model is a probability model corresponding to the second identification information; wherein the first identification information is used to indicate whether a node using DCM contains duplicate points, and the second identification information is used to indicate whether a leaf node of a multi-branch tree contains duplicate points; or, the first identification information is used to indicate whether a leaf node of a multi-branch tree contains duplicate points, and the second identification information is used to indicate whether a node using a direct coding mode contains duplicate points.

In an embodiment of the present application, the initial value of the probability model corresponding to the first identification information is set to be unequal to the first value (the first value corresponds to the situation where the probabilities of MPS and LPS are equal) and/or the probability model corresponding to the first identification information and the probability model corresponding to the second identification information are set to different probability models. The above setting method is more in line with the actual situation, thereby helping to improve the encoding and decoding efficiency of the point cloud.

The following is a more detailed example explanation of the embodiments of the present application in combination with Embodiment 1 and Embodiment 2. Embodiment 1 is described from the perspective of the DCM encoding and decoding process, and Embodiment 2 is described from the perspective of the encoding and decoding process of the leaf nodes of a multi-branch tree. In Embodiment 1, the first identification information is used to indicate whether the node using DCM contains duplicate points, and the second identification information is used to indicate whether the leaf nodes of the multi-branch tree contain duplicate points. In Embodiment 2, the first identification information is used to indicate whether the leaf nodes of the multi-branch tree contain duplicate points, and the second identification information is used to indicate whether the nodes using the direct coding mode contain duplicate points. It should be understood that the point cloud encoding and decoding process may require both DCM encoding and decoding and encoding and decoding of the leaf nodes of the multi-branch tree, so Embodiment 1 and Embodiment 2 can be combined with each other.

Example 1: DCM encoding and decoding

FIG9 is a flow chart of a decoding method provided in an embodiment of the present application. The method of FIG9 can be applied to a decoder. The decoder can be, for example, a decoder supporting G-PCC.

In step S910, the first identification information is decoded (such as entropy decoding) according to the first probability model. The first identification information is used to indicate whether the current node (the current node refers to the node using DCM) contains duplicate points. The first identification information can be, for example, a binary syntax element DupPointsCntGt0. The value of the first identification information may include a value and b value. Among them, the value a can be, for example, 1, and the value b can be, for example, 0. If the value of the first identification information is value a, it can be indicated that the current node contains duplicate points. If the value of the first identification information is value b, it can be indicated that the current node does not contain duplicate points.

The first probability model refers to a probability model for entropy decoding the first identification information. The MPS of the first probability model can be 0 (i.e., bin value 0), and the LPS can be 1 (i.e., bin value 1). The initial value of the first probability model may include the probability initial value of the MPS of the first probability model and/or the probability initial value of the LPS. In other words, the initial value of the first probability model can be used to represent the probability that the first probability model is 0 and/or 1.

In some implementations, the initial value of the first probability model is not equal to the first value. The first value corresponds to the case where the probabilities of MPS and LPS are equal. That is, if the initial value of the first probability model is the first value, the probabilities of MPS and LPS are equal. Since the probability that the first identification information takes a value of 0 is very high, setting the initial value of the first probability model to be not equal to the first value is more consistent with the actual situation, and can reduce the process of the decoder continuously updating the MPS interval width based on the renormalization operation, thereby improving decoding efficiency.

Exemplarily, the initial value of the first probability model can be set so that the probability of MPS is greater than or equal to 0.6, 0.7, 0.8 or 0.9. Optionally, the initial value of the first probability model can make the probability of MPS equal to 1. For example, the initial value of the first probability model can be equal to 1, which is 0xffff at 16-bit binary precision. Since the probability that the value of the first identification information is 0 is very high, setting the initial value of the first probability model to a value that makes the probability of MPS equal to 1 is more consistent with the actual situation, and can minimize the process of the decoder continuously updating the MPS interval width based on the renormalization operation, thereby improving decoding efficiency.

As mentioned above, the first probability model is used to decode the first identification information. In some implementations, in addition to being used to decode the first identification information, the first probability model can also be used to decode the second identification information. The second identification information is used to indicate whether the leaf node of a multi-branch tree (such as an octree, a quadtree, or a binary tree) contains duplicate points. The second identification information can be, for example, a binary syntax element DupPointsCntGt0L. The value of the second identification information may include a value and b value. Among them, the value a can be, for example, 1, and the value b can be, for example, 0. If the value of the second identification information is value a, it can be indicated that the multi-branch tree leaf node contains duplicate points. If the value of the second identification information is value b, it can be indicated that the multi-branch tree leaf node does not contain duplicate points.

However, the actual probability of a node using DCM containing duplicate points is different from the actual probability of a multi-tree tree leaf node containing duplicate points. Using the same probability model for entropy decoding is inconsistent with the actual situation and may reduce decoding efficiency. Therefore, in some other implementations, The second identification information may correspond to a second probability model (ie, the second identification information may be decoded using the second probability model). The probability models corresponding to the first identification information and the second identification information are set to different probability models, and the update processes of the two probability models do not affect each other, which helps to improve decoding efficiency.

In some implementations, before executing step S910, the third identification information may be decoded first. The third identification information is used to indicate whether the current node contains points with different geometric information. The third identification information may be, for example, a binary syntax element direct_point_cnt_eq2. The value of the third identification information may include a value and b value. Among them, the value a may be, for example, 1, and the value b may be, for example, 0. If the value of the third identification information is value a, it may indicate that the current node contains points with different geometric information (such as 2 points with different geometric information). If the value of the third identification information is value b, it may indicate that the current node contains only points with the same geometric information. For example, the current node includes only 1 point; or, the current node includes multiple points, and the geometric information of the multiple points is the same. If the third identification information indicates that the current node contains only points with the same geometric information, continue to execute step S910; otherwise, step S910 may not be executed.

Continuing to refer to FIG. 9 , in some implementations, the method of FIG. 9 may further include step S920 , that is, if the first identification information indicates that the current node contains duplicate points, determining the number of duplicate points contained in the current node.

For example, the fourth identification information can be decoded. The fourth identification information is used to indicate whether the current node contains 1 duplicate point. The fourth identification information can be, for example, a binary syntax element DupPointsCntGt1. The value of the fourth identification information may include a value and b value. Among them, the value a can be 1, for example, and the value b can be 0, for example. If the value of the fourth identification information is a, it can indicate that the current node contains multiple duplicate points (or more than 1 duplicate points). If the value of the fourth identification information is b, it can indicate that the current node contains 1 duplicate point. If the fourth identification information indicates that the current node contains 1 duplicate point, the number of duplicate points is determined to be 1.

Further, in some implementations, if the fourth identification information indicates that the current node includes multiple repeated points, the fifth identification information is decoded. The fifth identification information is used to indicate or determine the number of repeated points contained in the current node. The fifth identification information can be, for example, a binary syntax element DupPointsCntEg1. The fifth identification information can be, for example, based on 0-order exponential Columbus decoding. After parsing the value of the fifth identification information, the number of repeated points contained in the current node can be determined according to the value of the fifth identification information. For example, the sum of the value of the fifth identification information and 2 can be determined as the number of repeated points contained in the current node (such as DupPointsCntEg1+2).

As mentioned above, the current node is a node that uses DCM (if the current node does not use DCM, multitree decoding can be performed on the current node). Before executing step S910, it can be inferred whether the current node uses DCM (i.e., whether it has DCM decoding qualifications) based on the information of the parent node and/or the neighbor information of the parent node. The following is a detailed example of the inference method of whether the current node uses DCM.

In some implementations, it can be determined whether the first condition and/or the second condition are met based on the information of the parent node. The first condition may include that the number of occupied child nodes of the parent node is 1, and the number of occupied child nodes of the grandparent node of the current node (i.e., the parent node of the parent node) is less than or equal to 2 (i.e., the parent node and possibly another node). The second condition may include that the number of occupied child nodes of the parent node is 1, and the neighbor nodes of the parent node (the neighbor nodes mentioned here may refer to, for example, 6 neighbor nodes up, down, left, right, front and back) are not occupied. If the first condition and/or the second condition are met, it is determined that the current node uses DCM. For example, if the first condition is met, it is determined that the current node uses DCM. For another example, if the second condition is met, it is determined that the current node uses DCM. For another example, if the first condition and the second condition are met at the same time, it is determined that the current node uses DCM.

The following uses the first condition of "parent-based qualification" and the second condition of "6N qualification" as an example to illustrate in more detail whether the current node uses DCM decoding.

Whether the current node is eligible for DCM decoding may be determined based on the parent node of the current node itself or neighbor information of the parent node, which may be referred to as inference.

The inference of DCM decoding eligibility can be divided into the following two categories:

1) Parent-based eligibility: There is only one occupied child node at the parent node level (i.e. the current node), and the grandparent node (i.e. the parent node's parent node) has at most two occupied child nodes (i.e. the parent node and possibly one other node).

2) 6N qualification: There is only one occupied child node (i.e. the current node) at the parent node level, and no occupied neighbor nodes (the neighbors mentioned here refer to the 6 neighbor nodes above, below, left, right, front and back).

There are three possible inference methods for IDCM. Inference method 1 means: if the current node meets the parent-based qualification and 6N qualification, then DCM decoding is turned on. Inference method 2 means: if the current node meets the 6N qualification, then DCM decoding is turned on. Inference method 3 means: if the current node meets the parent-based qualification, then DCM decoding is turned on.

After decoding the first identification information, the value of the coding interval parameter can also be updated according to the value of the first identification information. The coding interval parameter may include a first parameter and/or a second parameter. The first parameter is used to indicate the starting point of the coding interval (the starting point of the coding interval of the arithmetic encoder). The first parameter can be represented by Low. The second parameter is used to indicate the width of the coding interval. The second parameter can be represented by Range. The initial value of the second parameter can be 510, represented by 9 bits. After updating the value of the coding interval parameter according to the value of the first identification information, the first probability model can be updated according to the value of the first identification information and the updated value of the coding interval parameter.

For example, first, determine whether the value of the first identification information is equal to LPS or MPS, so as to update Low and Range. Referring to FIG7 , if Bin=LPS, then Low=Low+R _MPS , Range=R _LPS ; if Bin=MPS, then Low remains unchanged and Range is equal to R _MPS . R _MPS represents the interval width corresponding to MPS, and R _LPS represents the interval width corresponding to LPS.

Then, the probability model is updated using Bin value, Low, and Range.

As Range is updated, the value of Range may be less than half of the initial interval length, that is, less than 256. At this time, a renormalization process is required. Figure 8 shows the process of the renormalization process. During the renormalization process, Low and Range will be shifted left (it should be noted that the "left shift operation on the parameter" mentioned in each embodiment of the present application refers to the left shift operation on the binary value of the parameter) until the value of Range is greater than half of the initial interval width. The left shift of Range is to expand the interval width of the arithmetic encoder, and the left shift of Low is to output bits, and the number of bits of the left shift of the two is the same.

The decoding method provided by the embodiment of the present application is described in detail above in conjunction with Figure 9. The encoding method provided by the embodiment of the present application is described in detail below in conjunction with Figure 10.

Fig. 10 is a flow chart of an encoding method provided in an embodiment of the present application. The method of Fig. 10 can be applied to a point cloud encoder. The point cloud encoder can be, for example, a decoder supporting G-PCC.

In step S1010, the first identification information is encoded (such as entropy coding) according to the first probability model. The first identification information is used to indicate whether the current node (the current node refers to the node using DCM) contains duplicate points. The first identification information can be, for example, a binary syntax element DupPointsCntGt0. The value of the first identification information may include a value and b value. Among them, the value a can be, for example, 1, and the value b can be, for example, 0. If the value of the first identification information is value a, it can be indicated that the current node contains duplicate points. If the value of the first identification information is value b, it can be indicated that the current node does not contain duplicate points.

The first probability model refers to a probability model for entropy encoding the first identification information. The MPS of the first probability model can be 0 (i.e., bin value 0), and the LPS can be 1 (i.e., bin value 1). The initial value of the first probability model may include the probability initial value of the MPS of the first probability model and/or the probability initial value of the LPS. In other words, the initial value of the first probability model can be used to represent the probability that the first probability model is 0 and/or 1.

In some implementations, the initial value of the first probability model is not equal to the first value. The first value corresponds to the case where the probabilities of MPS and LPS are equal. That is, if the initial value of the first probability model is the first value, the probabilities of MPS and LPS are equal. That is, in the first probability model, the initial interval width corresponding to MPS is greater than the initial interval width corresponding to LPS. Since the probability that the value of the first identification information is 0 is very high, setting the initial value of the first probability model to be not equal to the first value is more consistent with the actual situation, which can reduce the process of the encoder continuously updating the MPS interval width based on the renormalization operation, thereby improving the coding efficiency.

Exemplarily, the initial value of the first probability model can be set so that the probability of MPS is greater than or equal to 0.6, 0.7, 0.8 or 0.9. Optionally, the initial value of the first probability model can be set so that the probability of MPS is equal to 1. For example, the initial value of the first probability model can be equal to 1, which is 0xffff at 16-bit binary precision. Since the probability that the value of the first identification information is 0 is very high, setting the initial value of the first probability model to a value that makes the probability of MPS equal to 1 is more consistent with the actual situation, and can minimize the process of the encoder continuously updating the MPS interval width based on the renormalization operation, thereby improving the coding efficiency.

As mentioned above, the first probability model is used to encode the first identification information. In some implementations, in addition to being used to encode the first identification information, the first probability model can also be used to encode the second identification information. The second identification information is used to indicate whether the leaf node of a multi-branch tree (such as an octree, a quadtree, or a binary tree) contains duplicate points. The second identification information can be, for example, a binary syntax element DupPointsCntGt0L. The value of the second identification information may include a value and b value. Among them, the value a can be, for example, 1, and the value b can be, for example, 0. If the value of the second identification information is value a, it can be indicated that the multi-branch tree leaf node contains duplicate points. If the value of the second identification information is value b, it can be indicated that the multi-branch tree leaf node does not contain duplicate points.

However, the actual probability that a node using DCM contains duplicate points is different from the actual probability that a multi-tree tree leaf node contains duplicate points. Using the same probability model for entropy coding for both is inconsistent with the actual situation and may reduce coding efficiency. Therefore, in other implementations, the second identification information may correspond to the second probability model (i.e., the second identification information may be encoded using the second probability model). The probability models corresponding to the first identification information and the second identification information are set to different probability models, and the update processes of the two probability models do not affect each other, which helps to improve coding efficiency.

In some implementations, before executing step S1010, the third identification information may be encoded first. The third identification information is used to indicate whether the current node contains points with different geometric information. The third identification information may be, for example, a binary syntax element direct_point_cnt_eq2. The value of the third identification information may include a value and b value. Among them, the value a may be, for example, 1, and the value b may be, for example, 0. If the value of the third identification information is a, it may indicate that the current node contains points with different geometric information (such as 2 points with different geometric information). If the value of the third identification information is b, it may indicate that the current node contains only points with the same geometric information. For example, the current node includes only 1 point; or, the current node includes multiple points, and the geometric information of the multiple points is the same. If the third identification information indicates that the current node contains only points with the same geometric information, continue to execute step S1010; otherwise, step S1010 may not be executed.

Continuing to refer to FIG. 10 , in some implementations, the method of FIG. 10 may further include step S1020 , that is, if the first identification information indicates that the current node contains repeated points, encoding the number of repeated points contained in the current node.

For example, the fourth identification information may be encoded. The fourth identification information is used to indicate whether the current node contains one repeated point. The identification information may be, for example, a binary syntax element DupPointsCntGt1. The value of the fourth identification information may include a value and b value. The value a may be, for example, 1, and the value b may be, for example, 0. If the value of the fourth identification information is a value, it may indicate that the current node contains multiple duplicate points (or more than one duplicate point). If the value of the fourth identification information is b value, it may indicate that the current node contains one duplicate point. If the fourth identification information indicates that the current node contains one duplicate point, the number of duplicate points is 1.

Further, in some implementations, if the fourth identification information indicates that the current node includes multiple repeated points, the fifth identification information is encoded. The fifth identification information is used to indicate or determine the number of repeated points contained in the current node. The fifth identification information can be, for example, a binary syntax element DupPointsCntEg1. The fifth identification information can be, for example, based on a 0-order exponential Golomb coding. The number of repeated points contained in the current node can be equal to the sum of the value of the fifth identification information and 2 (such as DupPointsCntEg1+2).

As mentioned above, the current node is a node that uses DCM (if the current node does not use DCM, the current node can be multi-tree encoded). Before executing step S1010, it can be inferred whether the current node uses DCM (i.e., whether it has DCM encoding qualifications) based on the information of the parent node and/or the neighbor information of the parent node. The following is a detailed example of the inference method of whether the current node uses DCM.

In some implementations, it can be determined whether the first condition and/or the second condition are met based on the information of the parent node. The first condition may include that the number of occupied child nodes of the parent node is 1, and the number of occupied child nodes of the grandparent node of the current node (i.e., the parent node of the parent node) is less than or equal to 2 (i.e., the parent node and possibly another node). The second condition may include that the number of occupied child nodes of the parent node is 1, and the neighbor nodes of the parent node (the neighbor nodes mentioned here may refer to, for example, 6 neighbor nodes up, down, left, right, front and back) are not occupied. If the first condition and/or the second condition are met, it is determined that the current node uses DCM. For example, if the first condition is met, it is determined that the current node uses DCM. For another example, if the second condition is met, it is determined that the current node uses DCM. For another example, if the first condition and the second condition are met at the same time, it is determined that the current node uses DCM.

The following uses the first condition of "parent-based qualification" and the second condition of "6N qualification" as an example to illustrate in more detail whether the current node uses DCM encoding.

Whether the current node is eligible for DCM coding may be determined based on the parent node of the current node itself or the neighbor information of the parent node, which may be referred to as inference.

The inferences of DCM coding eligibility can be divided into the following two categories:

1) Parent-based eligibility: There is only one occupied child node at the parent node level (i.e. the current node), and the grandparent node (i.e. the parent node's parent node) has at most two occupied child nodes (i.e. the parent node and possibly one other node).

2) 6N qualification: There is only one occupied child node (i.e. the current node) at the parent node level, and no occupied neighbor nodes (the neighbors mentioned here refer to the 6 neighbor nodes above, below, left, right, front and back).

IDCM has three possible inference methods. Inference method 1 means: if the current node meets the parent-based qualifications and 6N qualifications, then turn on DCM encoding. Inference method 2 means: if the current node meets the 6N qualifications, then turn on DCM encoding. Inference method 3 means: if the current node meets the parent-based qualifications, then turn on DCM encoding.

After encoding the first identification information, the value of the coding interval parameter can also be updated according to the value of the first identification information. The coding interval parameter may include a first parameter and/or a second parameter. The first parameter is used to indicate the starting point of the coding interval (the starting point of the coding interval of the arithmetic encoder). The first parameter can be represented by Low. The second parameter is used to indicate the width of the coding interval. The second parameter can be represented by Range. The initial value of the second parameter can be 510, represented by 9 bits. After updating the value of the coding interval parameter according to the value of the first identification information, the first probability model can be updated according to the value of the first identification information and the updated value of the coding interval parameter.

For example, first, determine whether the value of the first identification information is equal to LPS or MPS, so as to update Low and Range. Referring to FIG7 , if Bin=LPS, then Low=Low+R _MPS , Range=R _LPS ; if Bin=MPS, then Low remains unchanged, and Range is equal to R _MPS . R _MPS represents the interval width corresponding to MPS, and R _LPS represents the interval width corresponding to LPS.

Then, the probability model is updated using Bin value, Low, and Range.

As Range is updated, the value of Range may be less than half of the initial interval length, that is, less than 256. At this time, a renormalization process is required. Figure 8 shows the process of the renormalization process. During the renormalization process, Low and Range are left shifted until the value of Range is greater than half of the initial interval width. The left shift of Range is to expand the interval width of the arithmetic encoder, and the left shift of Low is to output bits. The number of bits of the left shift of the two is the same.

Embodiment 1 is described in more detail below in conjunction with specific examples. It should be noted that the following examples are only intended to help those skilled in the art understand the embodiments of the present application, rather than to limit the embodiments of the present application to the specific numerical values or specific scenarios illustrated. Those skilled in the art can obviously make various equivalent modifications or changes based on the examples given, and such modifications or changes also fall within the scope of the embodiments of the present application.

The following are some of the syntax elements involved in this example.

If direct_point_cnt_eq2 in the above syntax element is 1, it means that the node includes 2 different points. If it is 0, it means that the node only contains points with the same geometric information (1 point, or multiple points with the same geometric information). direct_dup_point_cnt is used to identify the number of duplicate points. If direct_dup_point_cnt does not exist, the number of duplicate points in the node is inferred to be 0. If direct_dup_point_cnt exists, direct_dup_point_cnt identifies the number of duplicate points. The value of direct_dup_point_cnt plus 1 is the number of points in the node.

Encoding process:

As shown in Figure 5, in G-PCC, entropy coding of point coordinates in a node or subnode is called DCM. DCM is mainly used to process isolated nodes. The octree partitioning process uses neighbor context information for entropy coding. Unlike the octree partitioning process, DCM directly encodes the three component coordinates of the point.

For the node to be encoded, whether DCM is enabled is inferred from the information of the parent node and its neighbors, a process called IDCM. As shown in Figure 6, IDCM will first determine whether the current node is eligible for DCM encoding. If the current node is eligible for DCM encoding, the current node can be DCM encoded (of course, other judgment conditions can also be introduced, such as whether the number of points in the current node is less than a threshold (nb_points≤th), etc.). If the current node is not eligible for DCM encoding, the current node can be octree encoded.

Whether the current node is eligible for DCM coding may be determined based on the parent node of the current node itself or the neighbor information of the parent node, which may be referred to as inference.

The inferences of DCM coding eligibility can be divided into the following two categories:

1) Parent-based eligibility: There is only one occupied child node at the parent node level (i.e. the current node), and the grandparent node (i.e. the parent node's parent node) has at most two occupied child nodes (i.e. the parent node and possibly one other node).

2) 6N qualification: There is only one occupied child node (i.e. the current node) at the parent node level, and no occupied neighbor nodes (the neighbors mentioned here refer to the 6 neighbor nodes above, below, left, right, front and back).

In order to adapt to different compression efficiency and/or encoding time, IDCM has three possible inference methods. Inference method 1 means: if the current node meets the parent-based qualifications and 6N qualifications, DCM encoding is turned on. This inference method has the best compression effect, but the compression time is longer. Inference method 2 means: if the current node meets the 6N qualifications, DCM encoding is turned on. This inference method 2 relaxes the judgment conditions of DCM, and more nodes will enable DCM. Inference method 3 means: if the current node meets the parent-based qualifications, DCM encoding is turned on. This inference method uses more relaxed judgment conditions, but the compression efficiency is lower.

If the current node meets the DCM encoding qualification, a binary flag is first encoded to indicate whether the current node uses DCM. If the flag is 1, it means that the current node uses DCM; if the flag is 0, it means that the current node does not use DCM.

If the DCM mode is adopted, the following steps can be used to encode the current node.

First: Encode the binary syntax element direct_point_cnt_eq2. direct_point_cnt_eq2 identifies whether the current node contains two different points (i.e., the geometric information of the two points is different). If direct_point_cnt_eq2 is 1, it means that the current node contains two different points; if direct_point_cnt_eq2 is 0, it means that the current node only contains points with the same geometric information (it can contain only one point or multiple points with the same geometric information).

Second: If direct_point_cnt_eq2 is 0, encode the number of duplicate points direct_dup_point_cnt. First, encode the binary syntax element DupPointsCntGt0 to identify whether the current node contains duplicate points. If DupPointsCntGt0 is 1, it means that the current node contains duplicate points; if DupPointsCntGt0 is 0, it means that the current node does not contain duplicate points, and the number of duplicate points is 0. If DupPointsCntGt0 is 1, further encode the binary syntax element DupPointsCntGt1. DupPointsCntGt1 is used to identify whether the current node contains 1 duplicate point (that is, the geometric information of the two points is the same). If DupPointsCntGt1 is 1, it means that the current node contains more than 1 duplicate point; if DupPointsCntGt1 is 0, it means that the current node contains 1 duplicate point, and direct_dup_point_cnt=1. If DupPointsCntGt1 is 1, the syntax element DupPointsCntEg1 is further encoded. DupPointsCntEg1+2 is the number of repeated points, and direct_dup_point_cnt=DupPointsCntEg1+2.

When DupPointsCntGt0 is entropy encoded, the first probability model is used to entropy encode DupPointsCntGt0. The initial value of the first probability model (i.e., the probability that MPS is 0) can be set to 1. It is 0xffff at 16-bit binary precision. Since the probability that DupPointsCntGt0 takes the value of 0 is very high, setting the initial value of the first probability model to 1 is more consistent with the actual situation, which can minimize the process of the encoder continuously updating the MPS interval width based on the renormalization operation, thereby improving the encoding efficiency. In addition, DupPointsCntGt0 and DupPointsCntGt0L (used to identify whether a multi-tree tree leaf node contains duplicate points) use different probability models.

After encoding DupPointsCntGt0, the encoding interval parameter can also be updated according to the value of DupPointsCntGt0. Value. The coding interval parameter may include a first parameter and/or a second parameter. The first parameter is used to indicate the starting point of the coding interval (the coding interval starting point of the arithmetic encoder). The first parameter can be represented by Low. The second parameter is used to indicate the width of the coding interval. The second parameter can be represented by Range. The initial value of the second parameter can be 510, represented by 9 bits. After updating the value of the coding interval parameter according to the value of DupPointsCntGt0, the first probability model can be updated according to the value of DupPointsCntGt0 and the updated value of the coding interval parameter.

For example, first, determine whether the value of DupPointsCntGt0 is equal to LPS or MPS, so as to update Low and Range. Referring to FIG7 , if Bin=LPS, then Low=Low+R _MPS , Range=R _LPS ; if Bin=MPS, then Low remains unchanged, and Range is equal to R _MPS . R _MPS represents the interval width corresponding to MPS, and R _LPS represents the interval width corresponding to LPS.

Then, the probability model is updated using Bin value, Low, and Range.

As Range is updated, the value of Range may be less than half of the initial interval length, that is, less than 256. At this time, a renormalization process is required. Figure 8 shows the process of the renormalization process. During the renormalization process, Low and Range are left shifted until the value of Range is greater than half of the initial interval width. The left shift of Range is to expand the interval width of the arithmetic encoder, and the left shift of Low is to output bits. The number of bits of the left shift of the two is the same.

Decoding process:

For the node to be decoded, whether DCM is enabled is inferred from the information of the parent node and its neighbors, a process called IDCM. As shown in Figure 6, IDCM will first determine whether the current node is eligible for DCM decoding. If the current node is eligible for DCM decoding, DCM decoding can be performed on the current node (of course, other judgment conditions can also be introduced, such as whether the number of points in the current node is less than a threshold (nb_points≤th), etc.). If the current node is not eligible for DCM decoding, octree decoding can be performed on the current node.

Whether the current node is eligible for DCM decoding may be determined based on the parent node of the current node itself or neighbor information of the parent node, which may be referred to as inference.

The inference of DCM decoding eligibility can be divided into the following two categories:

1) Parent-based eligibility: There is only one occupied child node at the parent node level (i.e. the current node), and the grandparent node (i.e. the parent node's parent node) has at most two occupied child nodes (i.e. the parent node and possibly one other node).

2) 6N qualification: There is only one occupied child node (i.e. the current node) at the parent node level, and no occupied neighbor nodes (the neighbors mentioned here refer to the 6 neighbor nodes above, below, left, right, front and back).

In order to adapt to different compression efficiency and/or decoding time, IDCM has three possible inference methods. Inference method 1 means: if the current node meets the parent-based qualifications and 6N qualifications, DCM decoding is enabled. This inference method has the best compression effect, but the compression time is longer. Inference method 2 means: if the current node meets the 6N qualifications, DCM decoding is enabled. This inference method 2 relaxes the judgment conditions of DCM, and more nodes will enable DCM. Inference method 3 means: if the current node meets the parent-based qualifications, DCM decoding is enabled. This inference method uses more relaxed judgment conditions, but the compression efficiency is lower.

If the current node meets the DCM decoding qualification, a binary flag is first decoded to indicate whether the current node uses DCM. If the flag is 1, it means that the current node uses DCM; if the flag is 0, it means that the current node does not use DCM.

If the DCM mode is adopted, the following steps can be used to decode the current node.

First: Decode the binary syntax element direct_point_cnt_eq2. direct_point_cnt_eq2 identifies whether the current node contains two different points (i.e., the geometric information of the two points is different). If direct_point_cnt_eq2 is 1, it means that the current node contains two different points; if direct_point_cnt_eq2 is 0, it means that the current node only contains points with the same geometric information (it can contain only one point or multiple points with the same geometric information).

Second: If direct_point_cnt_eq2 is 0, decode the binary syntax element DupPointsCntGt0 to identify whether the current node contains duplicate points. If DupPointsCntGt0 is 1, it means that the current node contains duplicate points; if DupPointsCntGt0 is 0, it means that the current node does not contain duplicate points, that is, direct_dup_point_cnt = 0. If DupPointsCntGt0 is 1, further decode the binary syntax element DupPointsCntGt1. DupPointsCntGt1 is used to identify whether the current node contains 1 duplicate point (that is, the geometric information of the two points is the same). If DupPointsCntGt1 is 1, it means that the current node contains more than 1 duplicate points; if DupPointsCntGt1 is 0, it means that the current node contains 1 duplicate point, direct_dup_point_cnt = 1. If DupPointsCntGt1 is 1, further decode the syntax element DupPointsCntEg1. direct_dup_point_cnt=DupPointsCntEg1+2.

When entropy decoding DupPointsCntGt0, the first probability model can be used to entropy decode DupPointsCntGt0. The initial value of the first probability model (i.e., the probability that MPS is 0) can be set to 1. It is 0xffff at 16-bit binary precision. Since the probability that DupPointsCntGt0 takes the value of 0 is very high, setting the initial value of the first probability model to 1 is more consistent with the actual situation, which can minimize the process of the decoder continuously updating the MPS interval width based on the renormalization operation, thereby improving decoding efficiency. In addition, DupPointsCntGt0 and DupPointsCntGt0L (used to identify whether a multi-tree tree leaf node contains duplicate points) use different probability models.

After decoding DupPointsCntGt0, the encoding interval parameter can also be updated according to the value of DupPointsCntGt0. Value. The coding interval parameter may include a first parameter and/or a second parameter. The first parameter is used to indicate the starting point of the coding interval. The first parameter may be represented by Low. The second parameter is used to indicate the width of the coding interval. The second parameter may be represented by Range. The initial value of the second parameter may be 510, represented by 9 bits. After updating the value of the coding interval parameter according to the value of DupPointsCntGt0, the first probability model may be updated according to the value of DupPointsCntGt0 and the updated value of the coding interval parameter.

For example, first, determine whether the value of DupPointsCntGt0 is equal to LPS or MPS, so as to update Low and Range. Referring to FIG7 , if Bin=LPS, then Low=Low+R _MPS , Range=R _LPS ; if Bin=MPS, then Low remains unchanged, and Range is equal to R _MPS . R _MPS represents the interval width corresponding to MPS, and R _LPS represents the interval width corresponding to LPS.

Then, the probability model is updated using Bin value, Low, and Range.

As Range is updated, the value of Range may be less than half of the initial interval length, that is, less than 256. At this time, a renormalization process is required. Figure 8 shows the process of the renormalization process. During the renormalization process, Low and Range are left shifted until the value of Range is greater than half of the initial interval width. The left shift of Range is to expand the interval width, and the left shift of Low is to output bits. The number of bits of the left shift of the two is the same.

This technical solution reasonably uses the probability model and the correct initial value to further improve the efficiency of G-PCC geometric coding. The bits per input point ratio (Bpip Ratio) under the condition of lossless compression of geometric information indicates: the percentage of the encoding rate of this technical solution to the encoding rate of related technologies without loss of point cloud quality. The lower the value of this parameter, the greater the bit rate saved by this technical solution. Table 1 shows the test results of this technical solution. It can be seen from Table 1 that this technical solution has made a significant improvement in bit rate saving compared with related technologies.

Table 1 Bpip Ratio of geometry-based solid content test model (GES-TM)-v3.0 lossless compression (CW condition)

Embodiment 2: Multi-tree leaf node encoding and decoding

FIG11 is a flow chart of a decoding method provided in an embodiment of the present application. The method of FIG11 can be applied to a point cloud decoder. The decoder can be, for example, a decoder supporting G-PCC.

In step S1110, the first identification information is decoded (such as entropy decoding) according to the first probability model. The first identification information is used to indicate whether the current node contains duplicate points. Different from the first embodiment, the current node mentioned here refers to the leaf node of a multi-branch tree (such as an octree, a quadtree, or a binary tree). The first identification information can be, for example, a binary syntax element DupPointsCntGt0L. The value of the first identification information may include a value and b value. Among them, the value a can be, for example, 1, and the value b can be, for example, 0. If the value of the first identification information is value a, it can be indicated that the current node contains duplicate points. If the value of the first identification information is value b, it can be indicated that the leaf node of the multi-branch tree does not contain duplicate points.

The first probability model refers to a probability model for entropy decoding the first identification information. The MPS of the first probability model can be 0 (i.e., bin value 0), and the LPS can be 1 (i.e., bin value 1). The initial value of the first probability model may include the probability initial value of the MPS of the first probability model and/or the probability initial value of the LPS. In other words, the initial value of the first probability model can be used to represent the probability that the first probability model is 0 and/or 1.

In some implementations, the initial value of the first probability model is not equal to the first value (the probability that the first value corresponds to the MPS and the LPS is equal). That is, in the first probability model, the initial interval width corresponding to the MPS is greater than the initial interval width corresponding to the LPS. Since the probability that the value of the first identification information is 0 is very high, setting the initial value of the first probability model to be not equal to the first value is more consistent with the actual situation, and can reduce the process of the decoder continuously updating the MPS interval width based on the renormalization operation, thereby improving decoding efficiency.

Exemplarily, the initial value of the first probability model may be set so that the probability of MPS is greater than or equal to 0.6, 0.7, 0.8 or 0.9. Optionally, the initial value of the first probability model (i.e., the probability that MPS is 0) may be set so that the probability of MPS is equal to 1. The initial value of the first probability model may be, for example, equal to 1, or 0xffff at 16-bit binary precision. Since the probability that the first identification information is 0 is not Therefore, setting the initial value of the first probability model to a value that makes the probability of MPS equal to 1 is more consistent with the actual situation and can minimize the process of the decoder continuously updating the MPS interval width based on the renormalization operation, thereby improving decoding efficiency.

As mentioned above, the first probability model is used to decode the first identification information. In some implementations, in addition to being used to decode the first identification information, the first probability model can also be used to decode the second identification information. The second identification information is used to indicate whether the node using DCM contains duplicate points. The second identification information can be, for example, a binary syntax element DupPointsCntGt0. The value of the second identification information may include a value and b value. Among them, the value a can be, for example, 1, and the value b can be, for example, 0. If the value of the second identification information is value a, it can be indicated that the node using DCM contains duplicate points. If the value of the second identification information is value b, it can be indicated that the node using DCM does not contain duplicate points.

However, the actual probability that a node using DCM contains duplicate points is different from the actual probability that a multi-tree tree leaf node contains duplicate points. Using the same probability model for entropy decoding is inconsistent with the actual situation and may reduce decoding efficiency. Therefore, in other implementations, the second identification information may correspond to a second probability model (i.e., the second identification information may be decoded using a second probability model). The probability models corresponding to the second identification information and the first identification information are set to different probability models, and the update processes of the two probability models do not affect each other, which helps to improve decoding efficiency.

11 , in step S1120 , if the first identification information indicates that the current node contains duplicate points, the number of duplicate points contained in the current node is determined. For example, the number of duplicate points contained in the current node may be determined based on Exponential Golomb decoding.

After decoding the first identification information, the value of the coding interval parameter can also be updated according to the value of the first identification information. The coding interval parameter may include a first parameter and/or a second parameter. The first parameter is used to indicate the starting point of the coding interval (the starting point of the coding interval of the arithmetic encoder). The first parameter can be represented by Low. The second parameter is used to indicate the width of the coding interval. The second parameter can be represented by Range. The initial value of the second parameter can be 510, represented by 9 bits. After updating the value of the coding interval parameter according to the value of the first identification information, the first probability model can be updated according to the value of the first identification information and the updated value of the coding interval parameter.

For example, first, determine whether the value of the first identification information is equal to LPS or MPS, so as to update Low and Range. Referring to FIG7 , if Bin=LPS, then Low=Low+R _MPS , Range=R _LPS ; if Bin=MPS, then Low remains unchanged, and Range is equal to R _MPS . R _MPS represents the interval width corresponding to MPS, and R _LPS represents the interval width corresponding to LPS.

Then, the probability model is updated using Bin value, Low, and Range.

As Range is updated, the value of Range may be less than half of the initial interval length, that is, less than 256. At this time, a renormalization process is required. Figure 8 shows the process of the renormalization process. During the renormalization process, Low and Range are left shifted until the value of Range is greater than half of the initial interval width. The left shift of Range is to expand the interval width of the arithmetic encoder, and the left shift of Low is to output bits. The number of bits of the left shift of the two is the same.

The decoding method provided by the embodiment of the present application is described in detail above in conjunction with Figure 11. The encoding method provided by the embodiment of the present application is described in detail below in conjunction with Figure 12.

FIG12 is a flow chart of an encoding method provided in an embodiment of the present application. The method of FIG12 can be applied to a point cloud encoder. The point cloud encoder can be, for example, an encoder supporting G-PCC.

In step S1210, the first identification information is encoded (such as entropy coding) according to the first probability model. The first identification information is used to indicate whether the current node contains duplicate points. Different from the first embodiment, the current node mentioned here refers to the leaf node of a multi-branch tree (such as an octree, a quadtree, or a binary tree). The first identification information can be, for example, a binary syntax element DupPointsCntGt0L. The value of the first identification information may include a value and b value. Among them, the value a can be, for example, 1, and the value b can be, for example, 0. If the value of the first identification information is value a, it can be indicated that the current node contains duplicate points. If the value of the first identification information is value b, it can be indicated that the leaf node of the multi-branch tree does not contain duplicate points.

The first probability model refers to a probability model for entropy encoding the first identification information. The MPS of the first probability model can be 0 (i.e., bin value 0), and the LPS can be 1 (i.e., bin value 1). The initial value of the first probability model may include the probability initial value of the MPS of the first probability model and/or the probability initial value of the LPS. In other words, the initial value of the first probability model can be used to represent the probability that the first probability model is 0 and/or 1.

In some implementations, the initial value of the first probability model is not equal to the first value (the probability that the first value corresponds to the MPS and the LPS is equal). That is, in the first probability model, the initial interval width corresponding to the MPS is greater than the initial interval width corresponding to the LPS. Since the probability that the value of the first identification information is 0 is very high, setting the initial value of the first probability model to be not equal to the first value is more consistent with the actual situation, and can reduce the process of the encoder continuously updating the MPS interval width based on the renormalization operation, thereby improving the encoding efficiency.

Exemplarily, the initial value of the first probability model can be set so that the probability of MPS is greater than or equal to 0.6, 0.7, 0.8 or 0.9. Optionally, the initial value of the first probability model (i.e., the probability that MPS is 0) can be set so that the probability of MPS is equal to 1. The initial value of the first probability model can be equal to 1, for example, or 0xffff at 16-bit binary precision. Since the probability that the value of the first identification information is 0 is very high, setting the initial value of the first probability model to a value that makes the probability of MPS equal to 1 is more consistent with the actual situation, and can minimize the process of the decoder continuously updating the MPS interval width based on the renormalization operation, thereby improving decoding efficiency.

As mentioned above, the first probability model is used to encode the first identification information. In some implementations, in addition to encoding the first identification information In addition to encoding the information, the first probability model can also be used to encode the second identification information. The second identification information is used to indicate whether the node using DCM contains duplicate points. The second identification information can be, for example, a binary syntax element DupPointsCntGt0. The value of the second identification information may include a value and b value. Among them, the value a can be, for example, 1, and the value b can be, for example, 0. If the value of the second identification information is value a, it can be indicated that the node using DCM contains duplicate points. If the value of the second identification information is value b, it can be indicated that the node using DCM does not contain duplicate points.

However, the actual probability that a node using DCM contains duplicate points is different from the actual probability that a multi-tree tree leaf node contains duplicate points. Using the same probability model for entropy coding for both is inconsistent with the actual situation and may reduce coding efficiency. Therefore, in other implementations, the second identification information may correspond to the second probability model (i.e., the second identification information may be encoded using the second probability model). The probability models corresponding to the second identification information and the first identification information are set to different probability models, and the update processes of the two probability models do not affect each other, which helps to improve coding efficiency.

Continuing to refer to Fig. 12, in step S1220, if the first identification information indicates that the current node contains duplicate points, the number of duplicate points contained in the current node is encoded. For example, the number of duplicate points contained in the current node may be encoded based on the Exponential Golomb coding.

After encoding the first identification information, the value of the coding interval parameter can also be updated according to the value of the first identification information. The coding interval parameter may include a first parameter and/or a second parameter. The first parameter is used to indicate the starting point of the coding interval (the starting point of the coding interval of the arithmetic encoder). The first parameter can be represented by Low. The second parameter is used to indicate the width of the coding interval. The second parameter can be represented by Range. The initial value of the second parameter can be 510, represented by 9 bits. After updating the value of the coding interval parameter according to the value of the first identification information, the first probability model can be updated according to the value of the first identification information and the updated value of the coding interval parameter.

For example, first, determine whether the value of the first identification information is equal to LPS or MPS, so as to update Low and Range. Referring to FIG7 , if Bin=LPS, then Low=Low+R _MPS , Range=R _LPS ; if Bin=MPS, then Low remains unchanged, and Range is equal to R _MPS . R _MPS represents the interval width corresponding to MPS, and R _LPS represents the interval width corresponding to LPS.

Then, the probability model is updated using Bin value, Low, and Range.

As Range is updated, the value of Range may be less than half of the initial interval length, that is, less than 256. At this time, a renormalization process is required. Figure 8 shows the process of the renormalization process. During the renormalization process, Low and Range are left shifted until the value of Range is greater than half of the initial interval width. The left shift of Range is to expand the interval width of the arithmetic encoder, and the left shift of Low is to output bits. The number of bits of the left shift of the two is the same.

Embodiment 2 is described in more detail below in conjunction with specific examples. It should be noted that the following examples are only intended to help those skilled in the art understand the embodiments of the present application, rather than to limit the embodiments of the present application to the specific numerical values or specific scenarios illustrated. Those skilled in the art can obviously make various equivalent modifications or changes based on the examples given, and such modifications or changes also fall within the scope of the embodiments of the present application.

Encoding process:

In the multi-branch tree (such as octree) encoding of G-PCC, the multi-branch tree leaf nodes may contain duplicate points. The encoding method of the multi-branch tree leaf nodes includes the following steps. First, encode the binary syntax element DupPointsCntGt0L. DupPointsCntGt0L can be used to indicate whether the leaf node contains duplicate points. If DupPointsCntGt0L is 1, it means that the leaf node contains duplicate points; if DupPointsCntGt0L is 0, it means that the leaf node does not contain duplicate points. Then, when DupPointsCntGt0L is 1, the number of duplicate points is encoded using Exponential Columbus.

When encoding the binary syntax element DupPointsCntGt0L, the first probability model can be used to perform entropy coding on DupPointsCntGt0L. The initial value of the first probability model (i.e., the probability that MPS is 0) can be set to 1. It is 0xffff at 16-bit binary precision. Since the probability that DupPointsCntGt0L takes a value of 0 is very high, setting the initial value of the first probability model to 1 is more consistent with the actual situation, which can minimize the process of the encoder continuously updating the MPS interval width based on the renormalization operation, thereby improving the coding efficiency. In addition, DupPointsCntGt0L and DupPointsCntGt0 (used to identify whether a node using DCM contains duplicate points) use different probability models.

After encoding DupPointsCntGt0L, the value of the coding interval parameter can also be updated according to the value of DupPointsCntGt0L. The coding interval parameter may include a first parameter and/or a second parameter. The first parameter is used to indicate the starting point of the coding interval (the coding interval starting point of the arithmetic encoder). The first parameter can be represented by Low. The second parameter is used to indicate the width of the coding interval. The second parameter can be represented by Range. The initial value of the second parameter can be 510, represented by 9 bits. After updating the value of the coding interval parameter according to the value of DupPointsCntGt0L, the first probability model can be updated according to the value of DupPointsCntGt0L and the updated value of the coding interval parameter.

For example, first, determine whether the value of DupPointsCntGt0L is equal to LPS or MPS, so as to update Low and Range. Referring to FIG7 , if Bin=LPS, then Low=Low+R _MPS , Range=R _LPS ; if Bin=MPS, then Low remains unchanged, and Range is equal to R _MPS . R _MPS represents the interval width corresponding to MPS, and R _LPS represents the interval width corresponding to LPS.

Then, the probability model is updated using Bin value, Low, and Range.

As Range is updated, the value of Range may be less than half of the initial interval length, that is, less than 256. At this time, a renormalization process is required. Figure 8 shows the process of the renormalization process. During the renormalization process, Low and Range are shifted left. The operation is repeated until the value of Range is greater than half of the initial interval width. The left shift of Range is to expand the interval width of the arithmetic encoder, and the left shift of Low is to output bits. The number of bits of the left shift of the two is the same.

Decoding process:

The decoding method of a multi-tree leaf node includes the following steps. First, decode the binary syntax element DupPointsCntGt0L. If DupPointsCntGt0L is 1, it means that the leaf node contains duplicate points; if DupPointsCntGt0L is 0, it means that the leaf node does not contain duplicate points. When DupPointsCntGt0L is 1, continue to use the exponential Columbus decoding to decode the number of duplicate points.

When encoding the binary syntax element DupPointsCntGt0L, the first probability model can be used to entropy decode DupPointsCntGt0L. The initial value of the first probability model (i.e., the probability that MPS is 0) can be set to 1. It is 0xffff at 16-bit binary precision. Since the probability that DupPointsCntGt0L takes a value of 0 is very high, setting the initial value of the first probability model to 1 is more consistent with the actual situation, which can minimize the process of the decoder continuously updating the MPS interval width based on the renormalization operation, thereby improving decoding efficiency. In addition, DupPointsCntGt0L and DupPointsCntGt0 (used to identify whether a node using DCM contains duplicate points) use different probability models.

After decoding DupPointsCntGt0L, the value of the coding interval parameter can also be updated according to the value of DupPointsCntGt0L. The coding interval parameter may include a first parameter and/or a second parameter. The first parameter is used to indicate the starting point of the coding interval. The first parameter can be represented by Low. The second parameter is used to indicate the width of the coding interval. The second parameter can be represented by Range. The initial value of the second parameter can be 510, represented by 9 bits. After updating the value of the coding interval parameter according to the value of DupPointsCntGt0L, the first probability model can be updated according to the value of DupPointsCntGt0L and the updated value of the coding interval parameter.

For example, first, determine whether the value of DupPointsCntGt0L is equal to LPS or MPS, so as to update Low and Range. Referring to FIG7 , if Bin=LPS, then Low=Low+R _MPS , Range=R _LPS ; if Bin=MPS, then Low remains unchanged, and Range is equal to R _MPS . R _MPS represents the interval width corresponding to MPS, and R _LPS represents the interval width corresponding to LPS.

Then, the probability model is updated using Bin value, Low, and Range.

As Range is updated, the value of Range may be less than half of the initial interval length, that is, less than 256. At this time, a renormalization process is required. Figure 8 shows the process of the renormalization process. During the renormalization process, Low and Range are left shifted until the value of Range is greater than half of the initial interval width. The left shift of Range is to expand the interval width, and the left shift of Low is to output bits. The number of bits of the left shift of the two is the same.

This technical solution reasonably uses the probability model and the correct initial value to further improve the G-PCC geometric coding efficiency. Bpip Ratio under lossless compression of geometric information means: the percentage of the encoding rate of this technical solution to the encoding rate of related technologies without loss of point cloud quality. The lower the value of this parameter, the greater the bit rate saved by this technical solution. Table 2 shows the test results of this technical solution. It can be seen from Table 2 that this technical solution has made a significant improvement in bit stream saving compared with related technologies.

Table 2 Bpip Ratio of GES-TM-v3.0 lossless compression (CW condition)

The method embodiment of the present application is described in detail above in conjunction with Figures 1 to 12, and the device embodiment of the present application is described in detail below in conjunction with Figures 13 to 16. It should be understood that the description of the method embodiment corresponds to the description of the device embodiment, so the part not described in detail can refer to the previous method embodiment.

Figure 13 is a flow chart of a point cloud decoder provided by an embodiment of the present application. The point cloud decoder 1300 shown in Figure 13 includes a decoding unit 1310 and a determination unit 1320. The decoding unit 1310 is configured to decode the first identification information according to a first probability model, and the first identification information is used to indicate whether the current node contains duplicate points; the determination unit 1320 is configured to determine the number of duplicate points contained in the current node if the first identification information indicates that the current node contains duplicate points; wherein the first probability model satisfies at least one of the following: the initial value of the first probability model is not equal to the first value, and the first value corresponds to: the probabilities of MPS and LPS are equal; the first probability model is different from the second probability model, and the second probability model is a probability model corresponding to the second identification information; wherein, The first identification information is used to indicate whether a node using a direct coding mode contains duplicate points, and the second identification information is used to indicate whether a leaf node of a multi-branch tree contains duplicate points; or, the first identification information is used to indicate whether a leaf node of a multi-branch tree contains duplicate points, and the second identification information is used to indicate whether a node using a direct coding mode contains duplicate points.

In some implementations, the initial value of the first probability model is set so that the probability of the MPS is greater than or equal to 0.8.

In some implementations, the initial value of the first probability model is set so that the probability of the MPS is equal to 1.

In some implementations, the MPS is 0 and the LPS is 1.

In some implementations, the first identification information is used to indicate whether a node using a direct coding mode contains a repeated point.

In some implementations, the decoding unit 1310 is configured to: decode third identification information, where the third identification information is used to indicate whether the current node contains points with different geometric information; if the third identification information indicates that the current node only contains points with the same geometric information, decode the first identification information according to the first probability model.

In some implementations, the determination unit 1320 is configured to: decode fourth identification information, where the fourth identification information is used to indicate whether the current node contains 1 repeated point; if the fourth identification information indicates that the current node contains 1 repeated point, determine that the number of repeated points is 1.

In some implementations, the determination unit 1320 is further configured to: if the fourth identification information indicates that the current node includes multiple repeated points, decode the fifth identification information, and the fifth identification information is used to determine the number of repeated points contained in the current node; determine the number of repeated points contained in the current node based on the value of the fifth identification information.

In some implementations, the determination unit 1320 is configured to: determine the sum of the value of the fifth identification information and 2 as the number of repeated points included in the current node.

In some implementations, the decoding the fifth identification information includes: decoding the fifth identification information based on a 0th-order Exponential Columbus.

In some implementations, the determination unit 1320 is further configured to: before decoding the first identification information according to the first probability model, determine whether the current node uses a direct encoding mode according to information of a parent node of the current node.

In some implementations, the determination unit 1320 is configured to: determine whether the first condition and/or the second condition is met based on the information of the parent node; if the first condition and/or the second condition is met, determine that the current node uses a direct coding mode; wherein the first condition includes that the number of occupied child nodes of the parent node is 1, and the number of occupied child nodes of the grandparent node of the current node is less than or equal to 2; wherein the second condition includes that the number of occupied child nodes of the parent node is 1, and none of the neighboring nodes of the parent node are occupied.

In some implementations, the decoding unit 1310 is further configured to: if the current node does not use the direct coding mode, perform multi-tree decoding on the current node.

In some implementations, the first identification information is used to indicate whether a leaf node of the multi-branch tree contains a duplicate point.

In some implementations, the determining unit 1320 is configured to: determine the number of repeated points contained in the current node based on Exponential Golomb decoding.

In some implementations, the point cloud decoder 1300 also includes: an updating unit, configured to: update the value of the coding interval parameter according to the value of the first identification information, the coding interval parameter includes a first parameter and/or a second parameter, the first parameter is used to indicate the starting point of the coding interval, and the second parameter is used to indicate the width of the coding interval; update the first probability model according to the value of the first identification information and the updated value of the coding interval parameter.

In some implementations, the update unit is configured to: if the value of the first identification information is equal to MPS, determine the probability of the MPS as the updated value of the second parameter; and/or if the value of the first identification information is equal to LPS, determine the sum of the value of the first parameter and the probability of MPS as the updated value of the first parameter, and determine the probability of the LPS as the updated value of the second parameter.

In some implementations, the updating unit is further configured to: if the updated value of the second parameter is less than half of the initial value of the second parameter, perform a left shift operation on the first parameter and the second parameter.

It is understandable that in the embodiments of the present application, a "unit" may be a part of a circuit, a part of a processor, a part of a program or software, etc., and of course, it may be a module, or it may be non-modular. Moreover, the components in the present embodiment may be integrated into a processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of a software functional module.

If the integrated unit is implemented in the form of a software function module and is not sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this embodiment is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) or a processor to perform all or part of the steps of the method described in this embodiment. The aforementioned storage media include: U disk, mobile hard disk, read only memory (ROM), random access memory (RAM), Any medium that can store program code, such as a disk or CD.

Therefore, an embodiment of the present application provides a computer-readable storage medium, which is applied to the point cloud decoder 1300. The computer-readable storage medium stores a computer program, and the computer program implements the aforementioned decoding method when executed by a processor.

Based on the composition of the above-mentioned point cloud decoder and the computer-readable storage medium, refer to Figure 14, which shows a specific hardware structure diagram of the point cloud decoder provided in an embodiment of the present application. As shown in Figure 14, the point cloud decoder 1400 may include: a communication interface 1410, a memory 1420 and a processor 1430; each component is coupled together through a bus system 1440. It can be understood that the bus system 1440 is used to achieve connection and communication between these components. In addition to the data bus, the bus system 1440 also includes a power bus, a control bus and a status signal bus. However, for the sake of clarity, various buses are labeled as bus system 1440 in Figure 14. Among them,

The communication interface 1410 is used to receive and send signals when sending and receiving information with other external network elements.

The memory 1420 is used to store computer programs.

The processor 1430 is configured to, when running the computer program, execute:

Decoding first identification information according to a first probability model, where the first identification information is used to indicate whether the current node contains a duplicate point;

If the first identification information indicates that the current node includes duplicate points, determining the number of duplicate points included in the current node;

The first probability model satisfies at least one of the following:

The initial value of the first probability model is not equal to a first value, and the first value corresponds to: the probability of MPS and LPS being equal;

The first probability model is different from the second probability model, and the second probability model is a probability model corresponding to the second identification information; wherein the first identification information is used to indicate whether a node using a direct coding mode contains repeated points, and the second identification information is used to indicate whether a leaf node of a multi-branch tree contains repeated points; or, the first identification information is used to indicate whether a leaf node of a multi-branch tree contains repeated points, and the second identification information is used to indicate whether a node using a direct coding mode contains repeated points.

It can be understood that the memory 1420 in the embodiment of the present application can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory can be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of RAM are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), and direct RAM bus RAM (DRRAM). The memory 1420 of the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

The processor 1430 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit in the processor 1430 or the instruction in the form of software. The above processor 1430 may be a general processor, a digital signal processor (Digital Signal Processor, DSP), an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field programmable gate array (Field Programmable Gate Array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiments of the present application can be directly embodied as a hardware decoding processor to execute, or the hardware and software modules in the decoding processor can be executed. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 1420, and the processor 1430 reads the information in the memory 1420 and completes the steps of the above method in combination with its hardware.

It is understood that the embodiments described in this application can be implemented in hardware, software, firmware, middleware, microcode or a combination thereof. For hardware implementation, the processing unit can be implemented in one or more application specific integrated circuits (Application Specific Integrated Circuits, ASIC), digital signal processors (Digital Signal Processing, DSP), digital signal processing devices (DSP Device, DSPD), programmable logic devices (Programmable Logic Device, PLD), field programmable gate arrays (Field-Programmable Gate Array, FPGA), general processors, controllers, microcontrollers, microprocessors, other electronic units for performing the functions described in this application or a combination thereof. For software implementation, the technology described in this application can be implemented by a module (such as a process, function, etc.) that performs the functions described in this application. The software code can be stored in a memory and executed by a processor. The memory can be implemented in the processor or outside the processor.

Optionally, as another embodiment, the processor 1430 is further configured to execute the decoding method described in the above embodiment when running the computer program.

FIG15 is a schematic diagram of the structure of a point cloud encoder provided by an embodiment of the present application. The point cloud encoder 1500 of FIG15 includes a first encoding unit 1510 and a second encoding unit 1520. The first encoding unit 1510 is configured to encode the first identification information according to the first probability model. encoding, the first identification information is used to indicate whether the current node contains repeated points; the second encoding unit 1520 is configured to encode the number of repeated points contained in the current node if the first identification information indicates that the current node contains repeated points; wherein the first probability model satisfies at least one of the following: an initial value of the first probability model is not equal to a first value, and the first value corresponds to: the probabilities of MPS and LPS are equal; the first probability model is different from the second probability model, and the second probability model is a probability model corresponding to the second identification information; wherein the first identification information is used to indicate whether a node using a direct encoding mode contains repeated points, and the second identification information is used to indicate whether a leaf node of a multi-branch tree contains repeated points; or, the first identification information is used to indicate whether a leaf node of a multi-branch tree contains repeated points, and the second identification information is used to indicate whether a node using a direct encoding mode contains repeated points.

In some implementations, the initial value of the first probability model is set so that the probability of the MPS is greater than or equal to 0.8.

In some implementations, the initial value of the first probability model is set so that the probability of the MPS is equal to 1.

In some implementations, the MPS is 0 and the LPS is 1.

In some implementations, the first identification information is used to indicate whether a node using a direct coding mode contains a repeated point.

In some implementations, the first encoding unit 1510 is configured to: encode third identification information, where the third identification information is used to indicate whether the current node contains points with different geometric information; if the third identification information indicates that the current node only contains points with the same geometric information, encode the first identification information according to the first probability model.

In some implementations, the second encoding unit 1520 is configured to: encode fourth identification information, where the fourth identification information is used to indicate whether the current node contains 1 repeated point, and if the fourth identification information indicates that the current node contains 1 repeated point, the number of repeated points is 1.

In some implementations, the second encoding unit 1520 is further configured to: if the fourth identification information indicates that the current node includes multiple repeated points, encode the fifth identification information, and the number of repeated points is equal to the sum of the value of the fifth identification information and 2.

In some implementations, the second encoding unit 1520 is configured to: encode the fifth identification information based on a 0th-order exponential Golomb encoding.

In some implementations, the encoder 1500 further includes a first determination unit configured to determine whether the current node uses a direct encoding mode based on information of a parent node of the current node before encoding the first identification information based on the first probability model.

In some implementations, the determination unit is configured to: determine whether the first condition and/or the second condition is met based on the information of the parent node; if the first condition and/or the second condition is met, determine that the current node uses a direct coding mode; wherein the first condition includes that the number of occupied child nodes of the parent node is 1, and the number of occupied child nodes of the grandparent node of the current node is less than or equal to 2; wherein the second condition includes that the number of occupied child nodes of the parent node is 1, and none of the neighboring nodes of the parent node are occupied.

In some implementations, the encoder 1500 further includes a third encoding unit configured to: if the current node does not use a direct encoding mode, perform multi-tree encoding on the current node.

In some implementations, the first identification information is used to indicate whether a leaf node of the multi-branch tree contains a duplicate point.

In some implementations, the second encoding unit 1520 is configured to: encode the number of repeated points contained in the current node based on exponential Golomb coding.

In some implementations, the encoder 1500 also includes an updating unit, configured to: update the value of a coding interval parameter according to the value of the first identification information, the coding interval parameter including a first parameter and/or a second parameter, the first parameter being used to indicate the starting point of the coding interval, and the second parameter being used to indicate the width of the coding interval; update the first probability model according to the value of the first identification information and the updated value of the coding interval parameter.

In some implementations, the update unit is configured to: if the value of the first identification information is equal to the MPS, determine the probability of the MPS as the updated value of the second parameter; and/or, if the value of the first identification information is equal to the LPS, determine the sum of the value of the first parameter and the probability of the MPS as the updated value of the first parameter, and determine the probability of the LPS as the updated value of the second parameter.

In some implementations, the updating unit is further configured to: if the updated value of the second parameter is less than half of the initial value of the second parameter, perform a left shift operation on the first parameter and the second parameter.

It is understandable that in the embodiments of the present application, a "unit" may be a part of a circuit, a part of a processor, a part of a program or software, etc., and of course, it may be a module, or it may be non-modular. Moreover, the components in the present embodiment may be integrated into a processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of a software functional module.

If the integrated unit is implemented in the form of a software function module and is not sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this embodiment is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device) to perform operations. The aforementioned storage medium includes: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

Therefore, an embodiment of the present application provides a computer-readable storage medium, which is applied to the point cloud encoder 1500. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the encoding method in the aforementioned embodiment is implemented.

Based on the composition of the above-mentioned encoder 1500 and the computer-readable storage medium, refer to Figure 16, which shows a specific hardware structure diagram of the point cloud encoder provided in an embodiment of the present application. As shown in Figure 16, the point cloud encoder 1600 may include: a communication interface 1610, a memory 1620 and a processor 1630; each component is coupled together through a bus system 1640. It can be understood that the bus system 1640 is used to achieve connection and communication between these components. In addition to the data bus, the bus system 1640 also includes a power bus, a control bus and a status signal bus. However, for the sake of clarity, various buses are labeled as bus system 1640 in Figure 16. Among them,

The communication interface 1610 is used to receive and send signals when sending and receiving information with other external network elements.

The memory 1620 is used to store computer programs.

The processor 1630 is configured to, when running the computer program, execute:

Encoding first identification information according to a first probability model, where the first identification information is used to indicate whether the current node contains a duplicate point;

If the first identification information indicates that the current node includes repeated points, encoding the number of repeated points included in the current node;

The first probability model satisfies at least one of the following:

The initial value of the first probability model is not equal to a first value, and the first value corresponds to: the probability of MPS and LPS being equal;

The first probability model is different from the second probability model, and the second probability model is a probability model corresponding to the second identification information; wherein the first identification information is used to indicate whether a node using a direct coding mode contains repeated points, and the second identification information is used to indicate whether a leaf node of a multi-branch tree contains repeated points; or, the first identification information is used to indicate whether a leaf node of a multi-branch tree contains repeated points, and the second identification information is used to indicate whether a node using a direct coding mode contains repeated points.

It can be understood that the memory 1620 in the embodiment of the present application can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories. Among them, the non-volatile memory can be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory can be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of RAM are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), and direct RAM bus RAM (DRRAM). The memory 1620 of the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.

The processor 1630 may be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by the hardware integrated logic circuit in the processor 1630 or the instruction in the form of software. The above processor 1630 may be a general processor, a digital signal processor (Digital Signal Processor, DSP), an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field programmable gate array (Field Programmable Gate Array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The methods, steps and logic block diagrams disclosed in the embodiments of the present application can be implemented or executed. The general processor can be a microprocessor or the processor can also be any conventional processor, etc. The steps of the method disclosed in the embodiments of the present application can be directly embodied as a hardware decoding processor to execute, or the hardware and software modules in the decoding processor can be executed. The software module can be located in a mature storage medium in the field such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 1620, and the processor 1630 reads the information in the memory 1620 and completes the steps of the above method in combination with its hardware.

It is understood that the embodiments described in this application can be implemented in hardware, software, firmware, middleware, microcode or a combination thereof. For hardware implementation, the processing unit can be implemented in one or more application specific integrated circuits (Application Specific Integrated Circuits, ASIC), digital signal processors (Digital Signal Processing, DSP), digital signal processing devices (DSP Device, DSPD), programmable logic devices (Programmable Logic Device, PLD), field programmable gate arrays (Field-Programmable Gate Array, FPGA), general processors, controllers, microcontrollers, microprocessors, other electronic units for performing the functions described in this application or a combination thereof. For software implementation, the technology described in this application can be implemented by a module (such as a process, function, etc.) that performs the functions described in this application. The software code can be stored in a memory and executed by a processor. The memory can be implemented in the processor or outside the processor.

Optionally, as another embodiment, the processor 1630 is further configured to execute the aforementioned embodiment when running the computer program. The encoding method described.

It should be noted that, in this application, the terms "include", "comprises" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.

The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

The methods disclosed in several method embodiments provided in this application can be arbitrarily combined without conflict to obtain new method embodiments.

The features disclosed in several product embodiments provided in this application can be arbitrarily combined without conflict to obtain new product embodiments.

The features disclosed in several method or device embodiments provided in this application can be arbitrarily combined without conflict to obtain new method embodiments or device embodiments.

The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims (42)

A decoding method, applied to a point cloud decoder, comprising: Decoding first identification information according to a first probability model, where the first identification information is used to indicate whether the current node contains a duplicate point; If the first identification information indicates that the current node includes duplicate points, determining the number of duplicate points included in the current node; The first probability model satisfies at least one of the following: The initial value of the first probability model is not equal to a first value, and the first value corresponds to: the probability of the maximum probability symbol MPS and the minimum probability symbol LPS are equal; The first probability model is different from the second probability model, and the second probability model is a probability model corresponding to the second identification information; The first identification information is used to indicate whether a node using the direct coding mode contains a repeated point, and the second identification information is used to indicate whether a leaf node of the multi-branch tree contains a repeated point; or The first identification information is used to indicate whether a leaf node of the multi-branch tree contains a repeated point, and the second identification information is used to indicate whether a node using a direct coding mode contains a repeated point. The method according to claim 1, wherein the initial value of the first probability model is set so that the probability of the MPS is greater than or equal to 0.8. The method according to claim 1, wherein the initial value of the first probability model is set so that the probability of the MPS is equal to 1. The method according to claim 1, wherein the MPS is 0 and the LPS is 1. According to the method according to claim 1, the first identification information is used to indicate whether the node using the direct coding mode contains repeated points. The method according to claim 5, wherein decoding the first identification information according to the first probability model comprises: Decoding third identification information, where the third identification information is used to indicate whether the current node includes a point with different geometric information; If the third identification information indicates that the current node only includes points with the same geometric information, the first identification information is decoded according to the first probability model. The method according to claim 5, wherein the determining the number of repeated points contained in the current node comprises: Decoding fourth identification information, where the fourth identification information is used to indicate whether the current node includes one repeated point; If the fourth identification information indicates that the current node includes 1 repeated point, it is determined that the number of repeated points is 1. The method according to claim 7, wherein the determining the number of repeated points contained in the current node further comprises: If the fourth identification information indicates that the current node includes a plurality of repeated points, decoding the fifth identification information, where the fifth identification information is used to determine the number of repeated points included in the current node; The number of repeated points included in the current node is determined according to the value of the fifth identification information. The method according to claim 8, wherein the determining the number of repeated points contained in the current node according to the value of the fifth identification information comprises: The sum of the value of the fifth identification information and 2 is determined as the number of repeated points contained in the current node. The method according to claim 8 or 9, wherein decoding the fifth identification information comprises: The fifth identification information is decoded based on the 0th order Exponential Golomb decoding. The method according to claim 5, wherein, before decoding the first identification information according to the first probability model, the method further comprises: Determine whether the current node uses a direct coding mode according to information of the parent node of the current node. The method according to claim 11, wherein the determining whether the current node uses the direct coding mode according to the information of the parent node of the current node comprises: Determine whether the first condition and/or the second condition is satisfied according to the information of the parent node; If the first condition and/or the second condition is satisfied, determining that the current node uses a direct coding mode; The first condition includes that the number of occupied child nodes of the parent node is 1, and the number of occupied child nodes of the grandparent node of the current node is less than or equal to 2; The second condition includes that the number of occupied child nodes of the parent node is 1, and none of the neighboring nodes of the parent node are occupied. The method according to claim 11, wherein the method further comprises: If the current node does not use the direct coding mode, multitree decoding is performed on the current node. According to the method according to claim 1, the first identification information is used to indicate whether the leaf node of the multi-branch tree contains duplicate points. According to the method of claim 14, the determining the number of repeated points contained in the current node comprises: The number of duplicate points contained in the current node based on Exponential Columbus decoding. The method according to claim 1, wherein the method further comprises: Update the value of the coding interval parameter according to the value of the first identification information, the coding interval parameter includes a first parameter and/or a second parameter, the first parameter is used to indicate the starting point of the coding interval, and the second parameter is used to indicate the width of the coding interval; The first probability model is updated according to the value of the first identification information and the updated value of the coding interval parameter. The method according to claim 16, wherein updating the value of the coding interval parameter according to the value of the first identification information comprises: If the value of the first identification information is equal to the MPS, determining the probability of the MPS as the updated value of the second parameter; and/or If the value of the first identification information is equal to LPS, the sum of the value of the first parameter and the probability of MPS is determined as the updated value of the first parameter, and the probability of LPS is determined as the updated value of the second parameter. The method according to claim 17, wherein the method further comprises: If the updated value of the second parameter is less than half of the initial value of the second parameter, a left shift operation is performed on the first parameter and the second parameter. A coding method, applied to a point cloud encoder, comprising: Encoding first identification information according to a first probability model, where the first identification information is used to indicate whether the current node contains a duplicate point; If the first identification information indicates that the current node includes repeated points, encoding the number of repeated points included in the current node; The first probability model satisfies at least one of the following: The initial value of the first probability model is not equal to a first value, and the first value corresponds to: the probability of the maximum probability symbol MPS and the minimum probability symbol LPS are equal; The first probability model is different from the second probability model, and the second probability model is a probability model corresponding to the second identification information; The first identification information is used to indicate whether a node using the direct coding mode contains a repeated point, and the second identification information is used to indicate whether a leaf node of the multi-branch tree contains a repeated point; or, The first identification information is used to indicate whether a leaf node of the multi-branch tree contains a repeated point, and the second identification information is used to indicate whether a node using a direct coding mode contains a repeated point. The method according to claim 19, wherein the initial value of the first probability model is set so that the probability of the MPS is greater than or equal to 0.8. The method according to claim 19, wherein the initial value of the first probability model is set so that the probability of the MPS is equal to 1. The method according to claim 19, wherein the MPS is 0 and the LPS is 1. According to the method according to claim 19, the first identification information is used to indicate whether the node using the direct coding mode contains repeated points. The method according to claim 23, wherein encoding the first identification information according to the first probability model comprises: Encoding third identification information, where the third identification information is used to indicate whether the current node includes a point with different geometric information; If the third identification information indicates that the current node only includes points with the same geometric information, the first identification information is encoded according to the first probability model. The method according to claim 23, wherein encoding the number of repeated points contained in the current node comprises: The fourth identification information is encoded, where the fourth identification information is used to indicate whether the current node includes one repeated point. If the fourth identification information indicates that the current node includes one repeated point, the number of repeated points is 1. The method according to claim 25, wherein the encoding of the number of repeated points contained in the current node further comprises: If the fourth identification information indicates that the current node includes multiple repeated points, the fifth identification information is encoded, and the number of repeated points is equal to the sum of the value of the fifth identification information and 2. The method according to claim 26, wherein encoding the fifth identification information comprises: The fifth identification information is encoded based on the 0th order Exponential Golomb encoding. The method according to claim 23, wherein before encoding the first identification information according to the first probability model, the method further comprises: Determine whether the current node uses a direct coding mode according to information of the parent node of the current node. The method according to claim 28, wherein the determining whether the current node uses the direct encoding mode according to the information of the parent node of the current node comprises: Determine whether the first condition and/or the second condition is satisfied according to the information of the parent node; If the first condition and/or the second condition is satisfied, determining that the current node uses a direct coding mode; The first condition includes that the number of occupied child nodes of the parent node is 1, and the number of occupied child nodes of the grandparent node of the current node is less than or equal to 2; The second condition includes that the number of occupied child nodes of the parent node is 1, and none of the neighboring nodes of the parent node are occupied. The method according to claim 28, wherein the method further comprises: If the current node does not use the direct coding mode, multi-tree coding is performed on the current node. According to the method according to claim 19, the first identification information is used to indicate whether the leaf node of the multi-branch tree contains duplicate points. According to the method of claim 31, encoding the number of repeated points contained in the current node comprises: The number of duplicate points contained in the current node based on the Exponential Golomb code. The method according to claim 19, wherein the method further comprises: Update the value of the coding interval parameter according to the value of the first identification information, the coding interval parameter includes a first parameter and/or a second parameter, the first parameter is used to indicate the starting point of the coding interval, and the second parameter is used to indicate the width of the coding interval; The first probability model is updated according to the value of the first identification information and the updated value of the coding interval parameter. The method according to claim 33, wherein updating the value of the coding interval parameter according to the value of the first identification information comprises: If the value of the first identification information is equal to the MPS, determining the probability of the MPS as the updated value of the second parameter; and/or If the value of the first identification information is equal to LPS, the sum of the value of the first parameter and the probability of MPS is determined as the updated value of the first parameter, and the probability of LPS is determined as the updated value of the second parameter. The method according to claim 34, wherein the method further comprises: If the updated value of the second parameter is less than half of the initial value of the second parameter, a left shift operation is performed on the first parameter and the second parameter. A point cloud decoder, comprising: a decoding unit, configured to decode first identification information according to a first probability model, wherein the first identification information is used to indicate whether a current node contains a duplicate point; a determining unit, configured to determine the number of repeated points included in the current node if the first identification information indicates that the current node includes repeated points; The first probability model satisfies at least one of the following: The initial value of the first probability model is not equal to a first value, and the first value corresponds to: the probability of the maximum probability symbol MPS and the minimum probability symbol LPS are equal; The first probability model is different from the second probability model, and the second probability model is a probability model corresponding to the second identification information; The first identification information is used to indicate whether a node using the direct coding mode contains a repeated point, and the second identification information is used to indicate whether a leaf node of the multi-branch tree contains a repeated point; or The first identification information is used to indicate whether a leaf node of the multi-branch tree contains a repeated point, and the second identification information is used to indicate whether a node using a direct coding mode contains a repeated point. A point cloud decoder, comprising: Memory for storing computer programs; A processor, configured to execute the method according to any one of claims 1 to 18 when running the computer program. A point cloud encoder, comprising: A first encoding unit, configured to encode first identification information according to a first probability model, wherein the first identification information is used to indicate whether a current node contains a duplicate point; a second encoding unit, configured to encode the number of repeated points included in the current node if the first identification information indicates that the current node includes repeated points; The first probability model satisfies at least one of the following: The initial value of the first probability model is not equal to a first value, and the first value corresponds to: the probability of the maximum probability symbol MPS and the minimum probability symbol LPS are equal; The first probability model is different from the second probability model, and the second probability model is a probability model corresponding to the second identification information; The first identification information is used to indicate whether a node using the direct coding mode contains a repeated point, and the second identification information is used to indicate whether a leaf node of the multi-branch tree contains a repeated point; or The first identification information is used to indicate whether the leaf node of the multi-branch tree contains duplicate points, and the second identification information is used to indicate whether the leaf node of the multi-branch tree contains duplicate points. Whether the nodes in direct encoding mode contain duplicate points. A point cloud encoder, comprising: Memory for storing computer programs; A processor for executing the method according to any one of claims 19 to 35 when running the computer program. A non-volatile computer-readable storage medium storing a bit stream, wherein the bit stream is generated by an encoding method using an encoder, or the bit stream is decoded by a decoding method using a decoder, wherein the decoding method is the method described in any one of claims 1 to 18, and the encoding method is the method described in any one of claims 19 to 35. A code stream, comprising a code stream generated by the method according to any one of claims 19 to 35. A computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed, the method according to any one of claims 1 to 18 or 19 to 35 is implemented.