Point Cloud Geometry Coding with Cylindrical Regularization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing point cloud coding technologies based on prediction trees fail to fully reflect the spatial correlation of point clouds, leading to low prediction accuracy and coding efficiency due to non-uniform data distribution caused by noise and device jitter, necessitating additional information coding that reduces efficiency.
Innovation Solution
A method involving regularization preprocessing of point cloud data into a regularized structure using geometric distortion measures to ensure uniform distribution, followed by predictive coding based on lidar calibration information, which enhances spatial correlation and reduces the need for auxiliary bitstreams.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If point cloud data is directly encoded without preprocessing, then encoding process is simple, but encoding efficiency is low and resource waste occurs
Solution Approach 1:
The patent applies preliminary action by performing point cloud data preprocessing before encoding. Specifically, the system performs voxelization to convert point cloud data into a voxel structure, followed by traversing the voxel structure to generate traversal sequences. This preliminary organization of data into structured formats enables more efficient subsequent encoding operations, reducing resource waste and improving overall encoding efficiency.
2Productivity
If point cloud data is organized into structured formats through preprocessing, then encoding efficiency improves, but processing time increases
Solution Approach 1:
The patent applies segmentation by dividing the point cloud data processing into distinct stages: voxelization stage, traversal stage, and encoding stage. The voxelization process segments space into discrete voxels, and the traversal process segments the voxel structure into traversable sequences. This segmentation allows each stage to be optimized independently and enables parallel processing where applicable, reducing overall processing time while maintaining encoding efficiency.
3Loss of time
If point cloud data is processed in unstructured format, then processing is faster, but resource waste increases
Solution Approach 1:
The patent applies parameter changes by transforming the data structure parameters during preprocessing. The system changes from unstructured point cloud coordinates to structured voxel indices and traversal sequences. This parameter transformation organizes the data in a way that reduces redundancy and improves encoding efficiency, minimizing resource waste while maintaining acceptable processing speeds through optimized data representation.
Data Source
Figure 1~3
Figure 4~6
Figure 7~8
AI summary
The present invention discloses a point cloud data preprocessing method, a point cloud geometry coding method and apparatus, and a point cloud geometry decoding method and apparatus. The preprocessing method includes: performing coordinate conversion on original point cloud data to obtain a representation of an original point cloud in a cylindrical coordinate system; unfolding the cylindrical coordinate system to obtain a two-dimensional structure; and performing regularization processing on the two-dimensional structure based on a geometric distortion measure to obtain a regularized structure. The coding method includes: performing predictive coding on preprocessed original point cloud data to obtain a geometric information bitstream. According to the point cloud geometry coding method provided in the present invention, an original point cloud is subjected to regularization preprocessing, so that the point cloud presents regularized distribution in horizontal and vertical directions, which better reflects a spatial correlation of the point cloud, and improves coding efficiency.