Point Cloud Geometry Encoding with Azimuthal Context Interpolation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing point cloud compression methods, such as those in the G-PCC Test Model, suffer from sub-optimal compression efficiency due to inefficient context selection in azimuthal coding modes, which leads to increased computational complexity and reduced encoding performance.
Innovation Solution
An improved azimuthal coding mode that interpolates angles φleft,d and φright,d based on inherited angles and offsets associated with the parent interval, avoiding the systematic use of arc tangent functions while maintaining flexibility in context selection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If arc tangent functions are systematically used to determine angles in azimuthal coding mode, then coding precision is maintained, but computational complexity increases
Solution Approach 1:
The patent changes the parameter representation by using interpolated angles derived from parent interval angles instead of systematically computing arc tangent functions. The angles φleft,d and φright,d are interpolated based on the parent interval angle and offset values, transforming the computational approach from direct trigonometric calculation to interpolation-based parameter derivation.
Solution Approach 2:
The patent performs preliminary computation of parent interval angles once, then reuses these angles for multiple child intervals through interpolation. This preliminary action eliminates the need to repeatedly compute arc tangent functions for each child interval, significantly reducing computational complexity while maintaining coding precision.
2Productivity
If context selection is inefficient in azimuthal coding mode, then encoding performance deteriorates, but compression efficiency is already sub-optimal
Solution Approach 1:
The patent implements feedback-based context selection where the interpolated angles φleft,d and φright,d are used to dynamically determine the context for binary arithmetic encoding. The context selection adapts based on the actual angle values and point distribution patterns, creating a feedback loop that optimizes both encoding performance and compression efficiency simultaneously.
Solution Approach 2:
The patent applies local quality by selecting different contexts based on local angular characteristics. Each interval's context is determined by its specific interpolated angles and local point distribution, rather than using a uniform context selection strategy. This localized approach improves both encoding performance and compression efficiency in different regions of the point cloud.
Data Source
AI summary
A method of encoding a point cloud into a bitstream of encoded point cloud data representing a physical object is provided. The method includes an azimuthal coding mode providing a series of bits for encoding a coordinate of a point of the point cloud. The method includes: dividing a parent interval, to which the point coordinate belongs to, into a left half interval and a right half interval; interpolating a left angle (φleft,d), associated with the left half interval, and a right angle (φright,d), associated with the right half interval, from at least one inherited angle associated with the parent interval; selecting a context depending on the left angle (φleft,d) and the right angle (φright,d); and context-adaptive binary entropy encoding a bit (bd), into the bitstream, based on the selected context, the coded bit (bd) indicating which of the two half intervals the point coordinate belongs to.