3D Point Cloud Inter Prediction for Lower Compression 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 compression.
Innovation Solution
An encoding method that selects inter prediction points based on three-dimensional points with specific angle relationships, using a prediction tree structure to reduce residuals and processing amounts, and a decoding method that follows similar principles to ensure accurate reconstruction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If point cloud compression is applied to reduce data amount, then transmission and storage efficiency is improved, but coding complexity increases
Solution Approach 1:
The point cloud data is divided into multiple processing units (e.g., VPCUs - Vertical Point Cloud Units) that can be independently encoded. This segmentation allows the compression process to handle data in manageable chunks, reducing overall coding complexity while achieving effective compression of the total data amount.
Solution Approach 2:
The patent introduces vertical dimension exploitation by creating VPCUs that process points across multiple slices in the vertical direction. By organizing points into vertical groups and processing them together, the method adds a vertical processing dimension that improves compression efficiency without proportionally increasing complexity.
2Measurement precision
If more reference points are used for inter prediction, then prediction accuracy is improved, but processing load increases
Solution Approach 1:
The patent applies different prediction strategies to different regions of the point cloud. Intra prediction is used for certain areas while inter prediction with reference points is used for others. This local differentiation allows the system to achieve good overall prediction accuracy without applying the most computationally intensive method uniformly across all data.
Solution Approach 2:
Instead of using all available reference points for inter prediction, the method selectively uses a limited number of reference points (typically one or two) that provide sufficient prediction accuracy. This partial action approach achieves adequate prediction quality while significantly reducing the processing load compared to using all possible reference points.
3Quantity of substance
If point cloud data is compressed for transmission, then bandwidth requirements are reduced, but decoding complexity increases
Solution Approach 1:
The encoding process performs preliminary organization of point cloud data into VPCUs and establishes reference point relationships before compression. This preliminary structuring of the data allows the decoder to efficiently reconstruct the point cloud without having to perform complex analysis during decoding, thereby reducing decoding complexity while maintaining effective compression.
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.