Point Cloud Entropy Coding Using Neighbor Occupancy Context
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current point cloud compression methods are inefficient and ineffective, particularly in handling large datasets from applications like autonomous vehicles and virtual reality, where rapid and accurate encoding and decoding of 3D data are crucial.
Innovation Solution
The method involves encoding and decoding point clouds using a tree structure with recursive splitting of sub-volumes, where the occupancy pattern is determined for each node based on neighboring nodes' occupancy data, and probability distributions are selected and updated to entropy encode and decode the data effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional point cloud compression methods are used, then the encoding and decoding process is simpler, but the compression efficiency is insufficient for large datasets
Solution Approach 1:
The patent changes the parameter of probability distribution selection by introducing occupancy pattern-based context modeling. Instead of using fixed or simple adaptive probability models, the system selects probability distributions based on the occupancy patterns of neighboring nodes, dynamically adapting the coding parameters to local geometric structures. This resolves the contradiction by improving compression efficiency through parameter adaptation while keeping the complexity increase manageable.
Solution Approach 2:
The patent applies local quality by using neighboring node occupancy data to select probability distributions for the current node. Each node is coded with a probability distribution tailored to its local geometric context rather than using a global or uniform model. This local adaptation improves compression efficiency for regions with different geometric characteristics while maintaining a structured approach to complexity management.
2Productivity
If probability distributions are updated adaptively based on neighboring nodes, then compression performance improves by 4-20%, but the coding complexity increases
Solution Approach 1:
The patent implements partial action by updating probability distributions selectively based on occupancy patterns rather than continuously updating for every node. The system uses neighboring node occupancy data to determine when and how to adapt probability distributions, applying the complexity increase only where it provides compression benefits. This resolves the contradiction by achieving compression performance improvement while limiting the overall coding complexity increase.
3Measurement precision
If occupancy patterns are determined based on neighboring nodes, then the encoding accuracy improves, but the processing time increases
Solution Approach 1:
The patent applies preliminary action by determining occupancy patterns of neighboring nodes before encoding the current node. The system performs neighbor occupancy analysis in advance to select appropriate probability distributions, preparing the coding context before actual encoding occurs. This resolves the contradiction by improving encoding accuracy through preliminary context analysis while organizing the processing flow to minimize overall processing time.
Data Source
AI summary
Methods and devices for encoding a point cloud. A current node associated with a sub-volume is split into further sub-volumes, each further sub-volume corresponding to a child node of the current node, and, at the encoder, an occupancy pattern is determined for the current node based on occupancy status of the child nodes. A probability distribution is selected from among a plurality of probability distributions based on occupancy data for a plurality of nodes neighbouring the current node. The encoder entropy encodes the occupancy pattern based on the selected probability distribution to produce encoded data for the bitstream and updates the selected probability distribution. The decoder makes the same selection based on occupancy data for neighbouring nodes and entropy decodes the bitstream to reconstruct the occupancy pattern.