Point Cloud Occupancy Coding via Distance Context Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current point cloud compression techniques, such as those in the Test Model 13 (TMC13) codec, are not flexible enough to accommodate varying characteristics of point clouds, leading to hindered compression efficiency and unsuitability for all application scenarios due to rigid octree partitioning and traversal orders.
Innovation Solution
The proposed method involves context modeling of occupancy coding using flexible tree partitioning, such as octree, quad-tree, and binary-tree structures, with hybrid traversal orders and distance-based context modeling to improve coding efficiency and accommodate different point cloud characteristics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If rigid octree partitioning and traversal orders are used, then the coding process is simple, but compression efficiency is hindered and adaptability to varying point cloud characteristics is reduced
Solution Approach 1:
The patent implements dynamic selection of traversal orders (depth-first, breadth-first, and hybrid variants) based on point cloud characteristics. The coding system adapts its partitioning and traversal strategy during encoding, switching between different traversal modes to optimize compression efficiency for different scene types, thereby resolving the contradiction between adaptability and complexity.
Solution Approach 2:
The patent changes key parameters including traversal order selection, partitioning depth, and context modeling parameters based on point cloud density and spatial distribution characteristics. By dynamically adjusting these parameters, the system achieves high adaptability to varying point cloud characteristics without requiring completely different coding structures.
2Productivity
If context modeling based on distance between objects is implemented, then coding efficiency is improved, but computational complexity increases
Solution Approach 1:
The patent segments the point cloud into multiple objects or regions and applies context modeling based on distance relationships between these segmented elements. By dividing the point cloud into manageable segments and applying distance-based context modeling only to relevant neighboring segments, the system improves coding efficiency while controlling computational complexity through localized processing.
Solution Approach 2:
The patent applies different context modeling strategies based on local distance relationships between objects. Instead of uniform processing, the system adapts the context modeling approach according to the local spatial distribution and distance characteristics of point cloud objects, improving coding efficiency where needed while avoiding unnecessary computation in uniform regions.
Data Source
AI summary
A method for coding information of a point cloud comprises obtaining the point cloud including a set of points in a three-dimensional space; partitioning the point cloud into a plurality of objects and generating occupancy information for each of the plurality of objects; and encoding the occupancy information by taking into account the distance between the plurality of objects.