3D Point Encoding With Angular References for Lower Processing Load
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
There is a demand for improving coding efficiency in three-dimensional data encoding and decoding processes, particularly in point cloud representation methods.
Innovation Solution
An encoding method that selects inter prediction points based on three-dimensional points with specific angular relationships, using a prediction tree structure to reduce residuals and processing amounts, and incorporates multiple frames as reference points for enhanced coding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional point cloud encoding methods are used, then data can be compressed and transmitted, but coding efficiency is insufficient and processing amounts are large
Solution Approach 1:
The point cloud data is divided into multiple processing units (e.g., tiles or patches) that can be encoded and processed independently. This segmentation allows parallel processing and reduces the computational burden on each processing unit, thereby improving coding efficiency and reducing processing time
Solution Approach 2:
The method performs preliminary organization of three-dimensional points by distance and angle before encoding. By pre-sorting and structuring the point cloud data into processing units with specific angular relationships, the encoding process can proceed more efficiently with reduced computational requirements
2Measurement precision
If more reference three-dimensional points are selected for inter prediction, then prediction accuracy improves, but processing complexity increases
Solution Approach 1:
The method selects reference three-dimensional points based on local angular relationships specific to each processing unit. By choosing reference points with specific angular relationships (e.g., second angle corresponding to first angle, third angle greater than second angle) rather than using all available points, the solution achieves accurate local prediction while keeping processing complexity manageable
Solution Approach 2:
The invention changes the parameter selection criteria for reference points from arbitrary or exhaustive selection to selection based on specific angular relationships. This parameter-based selection method maintains prediction accuracy by ensuring geometrically relevant reference points are chosen while significantly reducing the number of points that need to be processed
Data Source
AI summary
An encoding method for encoding three-dimensional points each having a position represented by a distance and an angle, the encoding method comprising: identifying three-dimensional points that belong to a second processing unit and have been encoded, for inter prediction of a first three-dimensional point belonging to a first processing unit; and selecting a reference three-dimensional point from the three-dimensional points identified to calculate an inter predicted value of the first three-dimensional point. The three-dimensional points identified include a second three-dimensional point and a third three-dimensional point, the second three-dimensional point having a second angle corresponding to a first angle of the first three-dimensional point, the third three-dimensional point having a third angle greater than the second angle.