Point Cloud Entropy Coding With Neighbor and Child Contexts
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for compressing point clouds are inefficient and do not effectively manage contexts, leading to suboptimal compression of large datasets in applications like autonomous vehicles and virtual reality, where precise geometry coding is challenging due to sparse data populations.
Innovation Solution
The proposed solution involves a computer-implemented method for encoding and decoding point clouds using context-adaptive binary entropy coding, where probabilities for entropy coding are selected based on occupancy data from neighboring and child nodes in a tree structure, allowing for context reduction operations to minimize the number of contexts required, thereby improving compression efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If context-adaptive binary entropy coding is used for point cloud compression, then compression efficiency is improved, but the number of contexts to be managed increases excessively
Solution Approach 1:
The patent segments the context management by dividing the 3D space into octree nodes and further into sub-volumes (cubes). Each node's occupancy pattern is coded independently using binary entropy coding with contexts derived from neighboring nodes. This segmentation allows the system to manage contexts locally rather than globally, reducing the overall complexity while maintaining compression efficiency through localized adaptivity.
Solution Approach 2:
The patent applies local quality by using occupancy data from neighboring nodes and child nodes to select probabilities for entropy coding each specific node. Instead of using a single global context model, the system adapts the coding parameters locally based on the specific spatial configuration and occupancy patterns of surrounding regions, thereby improving compression efficiency without requiring an excessive number of global contexts.
2Productivity
If occupancy patterns are coded using binary entropy coding with neighboring node contexts, then compression performance is improved, but coding complexity increases
Solution Approach 1:
The patent performs preliminary action by pre-defining the octree structure and determining occupancy patterns for all nodes before the actual entropy coding process. The neighboring node relationships and child node configurations are established in advance, allowing the coding process to simply lookup and apply the appropriate contexts without complex real-time calculations, thus improving compression performance while controlling coding complexity.
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
AI summary
Methods and devices for encoding or decoding a point cloud. A bit sequence signalling an occupancy pattern for sub-volumes of a volume is coded using entropy coding. For a current sub-volume, probabilities of respective entropy coders for entropy coding the occupancy pattern may be selected based on occupancy data for a plurality of neighbouring sub-volumes of the current sub-volume and on occupancy data for subdivisions of the neighbouring sub-volumes.