Point Cloud Entropy Coding with Neighbor-Based Occupancy Context
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for compressing point cloud data are inefficient and do not effectively exploit local geometric correlations, leading to suboptimal compression and decoding processes.
Innovation Solution
The method involves encoding and decoding point clouds using a tree-based structure where the probability distribution for entropy encoding is selected based on occupancy data from neighboring nodes, rather than relying on a single fixed distribution, and updating this distribution based on the occupancy patterns of nearby nodes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If existing point cloud compression methods are used, then the compression process is simple, but the compression efficiency is low and local geometric correlations are not effectively exploited
Solution Approach 1:
The patent performs preliminary organization of point cloud data into overlapping local regions before entropy coding. This preliminary spatial organization enables the subsequent entropy coding stage to exploit local geometric correlations more effectively, improving compression efficiency without proportionally increasing overall complexity
Solution Approach 2:
The patent applies different probability distributions for entropy coding based on the specific occupancy patterns of local regions. By adapting the coding strategy to local geometric characteristics rather than using a uniform approach, the system improves compression efficiency by exploiting local correlations where they exist
2Productivity
If adaptive context modeling based on neighboring nodes is implemented, then compression performance improves by 4-20%, but coding complexity increases
Solution Approach 1:
The patent changes the parameter being coded by using occupancy patterns of neighboring nodes to select appropriate probability distributions. This parameter adaptation allows the system to achieve 4-20% compression improvement by matching the coding model to the actual geometric characteristics of each region
Solution Approach 2:
The patent segments the point cloud into overlapping local regions and processes each region independently with its own context modeling. This segmentation approach enables parallel processing and limits the complexity increase to only the necessary areas while maintaining overall compression performance
Data Source
Figure 1~2
Figure 3~4
Figure 5~6
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.