CN-115720270-B - Point cloud encoding method, decoding method, point cloud encoding device and decoding device

CN115720270BCN 115720270 BCN115720270 BCN 115720270BCN-115720270-B

Abstract

The invention discloses a point cloud encoding method, a decoding method, point cloud encoding equipment and decoding equipment, wherein the point cloud encoding method comprises the steps of calculating the distance from a first vertex to a father node of the first vertex as a first distance, determining the nearest point from the first vertex in uncoded vertexes as a second vertex, calculating residual values of N coordinate components from the second vertex to the first vertex, encoding absolute values of the residual values of the N coordinate components from the second vertex to the first vertex, and encoding signs of the residual values of the second vertex according to the residual values of the second vertex and the first distance. The method and the device utilize the geometric position relation among the vertexes in the point cloud to encode the absolute value of the residual value and the feasible sign bit in the residual value, so that the encoding efficiency of the residual value is improved, and meanwhile, the encoding performance of the point cloud prediction tree is improved by optimizing the encoding efficiency of the residual value.

Inventors

ZHAO WENBO
WANG JING
LI GE
GAO WEN

Assignees

鹏城实验室

Dates

Publication Date: 20260508
Application Date: 20210824

Claims (10)

1. A point cloud encoding method, characterized in that the point cloud encoding method comprises: Calculating the distance from a first vertex to a parent node of the first vertex as a first distance; Determining a point closest to the first vertex from uncoded vertexes as a second vertex, and calculating residual values from the second vertex to the first vertex on N coordinate components; Encoding absolute values of residual values on N coordinate components from the second vertex to the first vertex; and encoding the sign of the residual value of the second vertex according to the residual value of the second vertex and the first distance.
2. The method according to claim 1, wherein the encoding the sign of the residual value of the second vertex according to the residual value of the second vertex and the first distance specifically includes: traversing the signs of the residual values to obtain M residual values; Obtaining M vertexes according to the M residual values and the coordinate values of the first vertexes; calculating the distances from the father node to the M vertexes as a second distance; determining the feasibility of the sign of the residual value corresponding to the M vertexes according to the second distance and the first distance; And encoding the sign of the residual value of the second vertex according to the feasibility of the sign of the residual value.
3. The point cloud encoding method according to claim 2, wherein a weighted sum of magnitudes of differences of two points on N coordinate components to the power T is calculated as the first distance or the second distance; or calculating the maximum value of the weighted values of the magnitudes of the difference values of the two points on the N coordinate components as the first distance or the second distance.
4. The point cloud encoding method according to claim 2, wherein the encoding the sign of the residual value of the second vertex according to the feasibility of the sign of the residual value specifically includes: according to the feasibility of the symbols of the residual values corresponding to the M vertexes, the symbols of all feasible P vertexes are corresponding to new numbers; and encoding a new number corresponding to the sign of the residual value of the second vertex.
5. The method of point cloud encoding according to claim 4, wherein the encoding the new number corresponding to the sign of the residual value of the second vertex specifically includes: encoding binary bits of numbers corresponding to the residual values from high to low; when the ith bit is coded, the ith bit is 1, the rest uncoded bits are 0, and the value of the decimal number formed at present is marked as D; if D > P, the value of the bit cannot be 1, coding is not needed, otherwise, context-based arithmetic coding is carried out on the ith bit.
6. The point cloud encoding method according to claim 2, wherein the encoding the sign of the residual value of the second vertex according to the feasibility of the sign of the residual value specifically includes: Calculating the maximum value of the total number of the infeasible symbols after encoding a certain symbol bit; And ordering the symbol bits from the maximum value of the total number of the symbols which are not feasible after the encoding to the small value, and encoding the symbol bits in the order from the large value to the small value.
7. The point cloud coding device is characterized by comprising a processor, a memory and a communication bus, wherein the memory is stored with a computer readable program which can be executed by the processor; The communication bus realizes connection communication between the processor and the memory; the processor, when executing the computer readable program, implements the steps of the point cloud encoding method as claimed in any of claims 1-6.
8. A point cloud decoding method, characterized in that the point cloud decoding method comprises: decoding the point cloud code stream to obtain absolute values of residual values on N coordinate components from the second vertex to the first vertex; Calculating the distance from a first vertex to a parent node of the first vertex as a first distance; decoding the code stream according to the absolute value of the residual value of the second vertex and the first distance to obtain a symbol of the residual value of the second vertex; and obtaining the coordinate value of the second vertex according to the absolute value of the residual value of the second vertex and the sign of the residual value of the second vertex.
9. The method of point cloud decoding according to claim 8, wherein decoding the code stream according to the absolute value of the residual value of the second vertex and the first distance to obtain the sign of the residual value of the second vertex specifically includes: traversing the signs of the residual values to obtain M residual values; Obtaining M vertexes according to the M residual values and the coordinate values of the first vertexes; calculating the distances from the father node to the M vertexes as a second distance; determining the feasibility of the sign of the residual value corresponding to the M vertexes according to the second distance and the first distance; And decoding the code stream according to the feasibility of the sign of the residual value to obtain the sign of the residual value of the second vertex.
10. The point cloud decoding device is characterized by comprising a processor, a memory and a communication bus, wherein the memory is stored with a computer readable program which can be executed by the processor; The communication bus realizes connection communication between the processor and the memory; The processor, when executing the computer readable program, implements the steps in the point cloud decoding method according to any of claims 8-9.

Description

Point cloud encoding method, decoding method, point cloud encoding device and decoding device Technical Field The present invention relates to the field of point cloud encoding and decoding technologies, and in particular, to a point cloud encoding method, a point cloud decoding method, a point cloud encoding device, and a point cloud decoding device. Background A point cloud is a set of irregularly distributed discrete points in space that represent the spatial structure and surface properties of a three-dimensional object or scene. Each point in the point cloud has at least three-dimensional position information, and may also have color, material or other information according to the application scene. Typically, each point in the point cloud has the same number of additional attributes. The point cloud comprises a large number of unordered vertexes, when the geometric positions of the vertexes of the point cloud are coded, a coding scheme called a prediction tree exists in the current coding standard, namely, the vertexes of the point cloud are connected into a tree by adding directed edges between the vertexes, so that the vertexes have a sequential relationship. The position of the current node is predicted by the geometric position of the leading vertex of each vertex. And calculating residual values of the predicted value and the true value, and coding the residual values to realize the coding of the point cloud geometric information. In the prior art, the geometric prediction coding considers 3 preamble nodes of the current node (the preamble nodes represent nodes in front of the current node), selects the optimal mode from four different prediction modes, however, when the method codes the residual value, the sign bit of the residual value is arithmetically coded, and then the absolute value of the residual value is arithmetically coded, and the process is not optimized, so that the number of bits required for coding the residual value is large. In addition, for the geometric coding scheme of a single prediction mode, the method selects an uncoded point closest to the current node each time as a point to be coded and calculates a residual value, so that the prediction mode is not needed to be coded and only the residual value is needed to be coded, but in the method, when the residual value is coded, the sign bit of the residual value is arithmetically coded first, then the absolute value of the residual value is arithmetically coded, and the process is not optimized, so that the number of bits needed for coding the residual value is more. Accordingly, the prior art is still in need of improvement and development. Disclosure of Invention The invention mainly aims to provide a point cloud encoding method, a point cloud decoding method, a point cloud encoding device and a point cloud decoding device, and aims to solve the problems that in the prior art, when a residual value is encoded, the symbol bit of the residual value is arithmetically encoded first, then the absolute value of the residual value is arithmetically encoded, the process is not optimized, the number of bits required for encoding the residual value is large, and the residual value encoding efficiency is low. In order to achieve the above object, the present invention provides a point cloud encoding method, which includes the following steps: Calculating the distance from a first vertex to a parent node of the first vertex as a first distance; Determining a point closest to the first vertex from uncoded vertexes as a second vertex, and calculating residual values from the second vertex to the first vertex on N coordinate components; Encoding absolute values of residual values on N coordinate components from the second vertex to the first vertex; and encoding the sign of the residual value of the second vertex according to the residual value of the second vertex and the first distance. Optionally, the method for encoding a point cloud, wherein the encoding the sign of the residual value of the second vertex according to the residual value of the second vertex and the first distance specifically includes: traversing the signs of the residual values to obtain M residual values; Obtaining M vertexes according to the M residual values and the coordinate values of the first vertexes; calculating the distances from the father node to the M vertexes as a second distance; determining the feasibility of the sign of the residual value corresponding to the M vertexes according to the second distance and the first distance; And encoding the sign of the residual value of the second vertex according to the feasibility of the sign of the residual value. Optionally, the point cloud encoding method, wherein a weighted sum of the magnitudes of differences of the two points on the N coordinate components to the power T is calculated as the first distance or the second distance; or calculating the maximum value of the weighted values of the magnitudes of