Point Cloud Attribute Prediction With Limited Reference Buffer
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing methods for point cloud encoding and decoding are inefficient, particularly in terms of attribute encoding and decoding, leading to poor performance and challenges in transmitting large volumes of point cloud data.
Innovation Solution
A method and apparatus for point cloud encoding and decoding that involves determining a maximum number of reference points for a prediction reference buffer and using these points to calculate attribute prediction values, along with a point cloud encoder and decoder that utilize a prediction reference buffer to enhance encoding efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a large number of reference points are stored in the prediction reference buffer to improve attribute prediction accuracy, then the prediction precision improves, but the device complexity and memory requirements increase
Solution Approach 1:
The patent applies partial action by determining a limited maximum number M of reference points to store in the prediction reference buffer, rather than storing all available reference points. This selective approach balances prediction accuracy with memory constraints, storing only the necessary number of reference points needed for effective attribute prediction while avoiding excessive memory usage.
2Measurement precision
If more reference points are used in attribute prediction, then the attribute encoding precision improves, but the loss of time for processing increases
Solution Approach 1:
The patent limits the number of reference points to a maximum value M that is determined based on buffer capacity constraints. This partial action principle ensures that processing time is reduced by not considering an excessive number of reference points, while still maintaining adequate prediction precision through the use of the optimized subset of M reference points.
3Measurement precision
If the maximum number of reference points M is increased to improve prediction performance, then the attribute decoding quality improves, but the ease of operation and system simplicity deteriorates
Solution Approach 1:
The patent determines an optimized maximum number M of reference points that balances decoding quality with system simplicity. By selecting a specific limited number M rather than using all possible reference points, the system achieves good decoding performance while maintaining operational simplicity and avoiding the complexity of managing large numbers of reference points.
Data Source
AI summary
The present application provides a point cloud encoding method. The method includes: determining a first parameter, where the first parameter is used to indicate a maximum number M of reference points bufferable in a prediction reference buffer, M being a positive integer; determining M reference points based on the first parameter, and storing the M reference points into the prediction reference buffer, and determining an attribute prediction value of a current point based on reference points included in the prediction reference buffer.