Edge Neighborhood Entropy Coding for Lossless Point Cloud Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The large data size of point clouds necessitates efficient compression techniques for storage and transmission, with lossy compression impacting visual quality and lossless compression being crucial for certain applications like autonomous driving.
Innovation Solution
A point cloud coding system using an occupancy tree with dynamic OBUF and TriSoup entropy coding to reduce occupancy configurations, ensuring efficient compression and decompression while maintaining visual fidelity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If lossless compression is used for point cloud data, then data integrity is preserved, but compression ratio is limited
Solution Approach 1:
The point cloud data is segmented into multiple octants using an occupancy tree structure, where each octant is independently processed. This segmentation allows the application of different coding strategies (lossless vs. lossy) to different regions, enabling overall compression while preserving integrity where needed.
Solution Approach 2:
The patent applies different compression qualities to different regions of the point cloud. Critical regions maintain lossless compression for data integrity, while less critical regions use lossy compression to achieve higher compression ratios, thus resolving the contradiction between integrity and compression ratio.
2Loss of substance
If point cloud data is compressed for transmission, then storage and transmission efficiency improve, but processing complexity increases
Solution Approach 1:
The occupancy tree structure and octant decomposition are performed in advance during encoding. This preliminary organization of data into a hierarchical structure simplifies subsequent decoding and processing operations, reducing processing complexity while maintaining compression efficiency.
Solution Approach 2:
The patent employs dynamic programming techniques in the entropy coding process, where context models are adaptively updated based on previously decoded symbols. This dynamic approach optimizes compression ratios without requiring excessively complex processing, as the adaptation follows systematic rules rather than exhaustive search.
3Loss of substance
If occupancy tree depth is increased for better compression, then compression ratio improves, but visual quality deteriorates
Solution Approach 1:
Instead of uniformly increasing occupancy tree depth throughout the entire point cloud, the patent applies deeper decomposition only to specific octants where it provides benefit. This partial application maintains visual quality in critical regions while achieving better compression in less sensitive regions.
Solution Approach 2:
The patent dynamically adjusts the occupancy tree depth parameter based on local characteristics of different octants. Regions with simple geometry use shallower trees to preserve quality, while complex regions tolerate deeper trees for improved compression, thus balancing compression ratio and visual quality through parameter adaptation.
Data Source
AI summary
One or more methods, apparatuses, computer-readable storage mediums, and systems for entropy coding vertex information of an edge in a voxelized space of a point cloud are disclosed. Symbols of a neighborhood configuration of a current edge may be determined based on one or more already coded edges. The already coded edges may be selected from a spatial topology of edges or its subset. A lookup table may be used to retrieve an index for a given occupancy configuration for a neighborhood of a current edge. The index may indicate an appropriate context or probability model. Based on the coding of the current edge, the lookup table may be updated.