Scalable Point Cloud Attribute Coding via Bit-Plane Iteration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Graph-based Point Cloud Compression (G-PCC) technologies face challenges in achieving scalable lossless or near-lossless coding of attributes, particularly in reconstructing data from lossy to lossless fidelity, and do not efficiently handle the coding of lifting coefficients.
Innovation Solution
The proposed solution involves modifying the current G-PCC lifting design to enable scalable coding of lifting coefficients by iterating over multiple bit-planes of transform coefficients, processing all points in the point cloud, and generating an embedded bitstream, which allows for flexible fidelity adjustment from lossy to lossless reconstruction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If current G-PCC lifting design is used for attribute coding, then coding efficiency is improved, but scalable lossless or near-lossless coding capability is insufficient
Solution Approach 1:
The patent introduces dynamic bit-plane iteration over lifting coefficients, allowing the coding system to adapt between lossy and lossless modes by controlling the number of bit-planes processed. This dynamic approach enables scalable fidelity reconstruction while maintaining coding efficiency through the existing lifting framework.
Solution Approach 2:
The patent changes the parameter of fidelity by iterating over multiple bit-planes of transform coefficients. By controlling the iteration depth and bit-plane precision, the system can adjust between lossy and lossless coding modes, achieving scalable reconstruction without fundamentally changing the lifting structure.
2Measurement precision
If lossless fidelity reconstruction is achieved, then data accuracy is improved, but data volume and processing complexity increase
Solution Approach 1:
The patent applies partial action by selectively processing only the necessary bit-planes of lifting coefficients based on the desired fidelity level. For lossless coding, all bit-planes are processed; for lossy coding, only significant bit-planes are processed, reducing data volume while maintaining acceptable accuracy.
Solution Approach 2:
The patent segments the attribute data into multiple bit-planes of lifting coefficients, allowing independent processing of each bit-plane. This segmentation enables progressive transmission and decoding, where lower bit-planes can be processed separately from higher bit-planes, optimizing the balance between accuracy and data volume.
3Adaptability or versatility
If multiple bit-planes of transform coefficients are processed, then scalable fidelity is achieved, but coding complexity increases
Solution Approach 1:
The patent performs preliminary action by pre-organizing the lifting coefficients into multiple bit-planes before the main coding process. This preliminary structuring simplifies the subsequent iteration process, as the bit-plane organization is already established and can be accessed sequentially without complex reorganization during coding.
Data Source
AI summary
A method of encoding video data corresponding to a point cloud by at least one processor, the method including obtaining a plurality of transform coefficients corresponding to attributes of the point cloud; and encoding the plurality of transform coefficients to generate an embedded bitstream, the encoding including iterating over a plurality of bit-planes of the plurality of transform coefficients to process all points in the point cloud.