Mesh Point Cloud Coding with Wavelet LoD Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing geometry-based point cloud compression (G-PCC) standards face challenges in handling a wide range of point cloud inputs due to complex syntax element representation and organization, leading to increased complexity and inefficiency across various applications.
Innovation Solution
A method and system for encoding and decoding meshes represented in a 1D array using mesh segmentation, decimation, wavelet transform, and zero-run length coding to improve flexibility and generality, compatible with G-PCC standards like AVS and MPEG.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If G-PCC standards use complex syntax element representation and organization to handle various point cloud inputs, then the adaptability to different point cloud formats is improved, but the device complexity and computational overhead increase
Solution Approach 1:
The patent segments the point cloud data into multiple octrees organized in a quadtree structure, where each node represents a spatial region. This hierarchical segmentation allows the system to handle diverse point cloud inputs efficiently by processing only relevant segments, reducing the complexity of syntax element representation while maintaining adaptability across different input formats.
2Quantity of substance
If point cloud compression is applied to reduce memory and bandwidth requirements, then the loss of information increases due to compression artifacts
Solution Approach 1:
The patent implements dynamic refinement where the point cloud data structure adapts its resolution based on spatial importance and viewer position. Critical regions maintain high fidelity with minimal compression, while less important regions use higher compression ratios. This dynamic approach minimizes information loss in essential areas while achieving overall compression of memory and bandwidth requirements.
Solution Approach 2:
Different compression strategies are applied to different spatial regions of the point cloud based on their local characteristics and importance. High-priority regions use lossless or near-lossless coding, while lower-priority regions use more aggressive compression. This local quality differentiation reduces overall information loss while achieving compression of the total data volume.
3Measurement precision
If higher resolution point cloud capture technology is used to improve measurement precision, then the quantity of data increases requiring more memory and bandwidth
Solution Approach 1:
The patent employs a nested octree structure where multiple levels of detail are embedded within each other. Each octree level represents a different resolution, with finer levels nested within coarser levels. This allows the system to store and transmit only the necessary detail level for each region, maintaining high measurement precision where needed while reducing overall data size by eliminating redundant high-resolution information in less critical areas.
Data Source
AI summary
A method for encoding a mesh that is represented in a one-dimensional (1D) array is provided. The method includes: applying mesh segmentation to the mesh to generate a segmented mesh; performing mesh decimation to the segmented mesh to generate a subdivided mesh; calculating a set of mesh displacements for at least one level-of-detail (LoD) based on the segmented mesh and the subdivided mesh using a coding mode of a plurality of coding modes; applying a wavelet transform to the set of mesh displacements to generate a plurality of wavelet-transform coefficients; quantizing the plurality of wavelet-transform coefficients to generate a plurality of quantized wavelet-transform coefficients; and encoding the plurality of quantized wavelet-transform coefficients to a zero-run length code.