Point Cloud Transform Coefficient Coding for Random Access
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing point cloud encoding and decoding methods result in high memory overheads and inefficient spatial random access due to uniform encoding of transform coefficients, leading to uncontrollable processes and low efficiency.
Innovation Solution
A point cloud encoding method that limits the number of groups to be encoded based on encoding limit information, performs prediction processing to obtain attribute residuals, and sorts transform coefficients to optimize encoding and decoding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional video coding tools are used for point cloud data, then the coding structure can be maintained, but the coding efficiency is insufficient and large bitrates are required
Solution Approach 1:
The patent transforms point cloud data from spatial coordinates (x, y, z) into depth map representations, changing the parameter representation from 3D spatial coordinates to 2D depth values. This parameter transformation enables the use of conventional video coding tools while achieving significantly improved coding efficiency and reduced bitrate for point cloud data
Solution Approach 2:
The patent creates depth map copies from point cloud data by projecting 3D points onto 2D planes from different viewpoints. These depth map copies serve as intermediaries that can be efficiently coded using existing video coding standards, avoiding the need to directly code the original complex 3D point cloud data
2Productivity
If point cloud data is directly coded without transformation, then the original 3D information is preserved, but the coding complexity increases and efficiency decreases
Solution Approach 1:
The patent changes the parameter representation from raw 3D point cloud coordinates to depth map values, transforming the data structure into a form that is more amenable to conventional video coding processes. This parameter change simplifies the coding complexity while maintaining coding efficiency
Solution Approach 2:
The patent segments the point cloud data into multiple depth map representations from different viewpoints and depth ranges. This segmentation breaks down the complex 3D coding problem into simpler 2D depth map coding tasks that can be handled by existing video coding tools
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
The embodiments of the present application relate to the field of big data. Disclosed are a point cloud coding method and a point cloud decoding method. The point cloud coding method comprises: acquiring coding limit information, and on the basis of the coding limit information, acquiring a sequence of transformation coefficients to be coded of point cloud points to be coded, said point cloud points being included in a point cloud group to be coded, wherein the coding limit information comprises a limit threshold value of the number of groups, the number of point cloud groups to be coded is less than or equal to the limit threshold value of the number of groups, and said sequence is obtained by means of sorting transformation coefficients of said point cloud points (301); and coding said sequence, so as to obtain a point cloud group code stream corresponding to the point cloud group to be coded (302).