Point Cloud Direct Coding for Faster Multi-Point Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing point cloud compression techniques, particularly those using tree-based structures, face inefficiencies in handling isolated points and suffer from high computational burden and bandwidth/memory costs due to recursive splitting and signaling, which negatively impact compression performance.
Innovation Solution
The implementation of Direct Coding Mode (DCM) with Inferred Direct Coding Mode (IDCM) that conditionally applies direct coding based on a threshold number of points within a sub-volume, and the use of pairwise encoding for unordered points to improve compression efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If tree-based structures with recursive splitting are used for point cloud compression, then compression structure is established, but computational burden and encoding/decoding time increase significantly
Solution Approach 1:
The patent extracts isolated points from the tree-based structure processing and applies direct coding mode specifically to these points. By identifying and separating points that meet the isolation criterion (number of points in sub-volume is less than or equal to a threshold), the system avoids recursive splitting for these points, thereby reducing computational burden while maintaining compression structure for other points.
Solution Approach 2:
The patent segments the point cloud processing into two distinct paths: direct coding mode for isolated points and tree-based structure for non-isolated points. This segmentation allows each processing path to be optimized independently, with direct coding providing faster processing for isolated points and tree-based structure providing efficient compression for clustered points.
2Productivity
If direct coding mode is applied to isolated points, then encoding and decoding time is reduced, but compression performance may deteriorate without optimized coordinate encoding
Solution Approach 1:
The patent merges pairwise encoding with direct coding mode by encoding pairs of coordinate values (x1, x2), (y1, y2), (z1, z2) together rather than individually. This merging exploits the spatial correlation between nearby isolated points, allowing the encoder to represent coordinate pairs more efficiently and maintain compression performance while achieving faster coding speed.
Solution Approach 2:
The patent changes the encoding parameter from individual coordinate values to paired coordinate values. By encoding coordinates in pairs and using the relationship between paired points, the system achieves more efficient representation that maintains compression performance. The parameter change from single-point encoding to two-point paired encoding enables better exploitation of spatial correlations.
3Loss of information
If pairwise encoding is used for unordered points, then compression efficiency is improved, but complexity of determining point pairs increases
Solution Approach 1:
The patent applies self-service by using the coordinate values themselves to determine pairing. Points are paired based on their spatial coordinates without requiring external classification or complex algorithms. The system automatically identifies and pairs points based on their position information, eliminating the need for separate classification steps and reducing overall complexity.
Data Source
AI summary
Methods and devices for coding point clouds using direct coding mode to code coordinates of a point within a sub-volume associated with a current node instead of a pattern of occupancy for child nodes. When direct coding is applied to two or more points in the sub-volume, the points are ordered based on one of their respective coordinate values in binary and pairwise coding of those coordinate values is carried out on a bit-by-bit basis starting from the most significant bit position. For example, the coordinate values relate to a direction in the Cartesian coordinate system, the location of the first point is represented by an x-direction coordinate value, a y-direction coordinate value and a z-direction coordinate value. The pairwise coding includes coding whether the bits are the same and, if so, coding the bit value.