Subdivision-Improved Displacement Generation for Dynamic Mesh Compression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mesh compression standards fail to efficiently compress dynamic meshes with time-varying connectivity information and attribute maps, which are essential for realistic 3D content representation and transmission.
Innovation Solution
The method involves subdividing a polygon mesh to generate sub-vertices, determining normal vectors for edge vertices, moving sub-vertices along these normal vectors, and calculating displacements between moved sub-vertices and reference vertices, all while generating a bitstream that encodes these displacements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If existing mesh compression standards are used to compress dynamic meshes with time-varying connectivity information and attribute maps, then compression is achieved, but the signaling cost for displacement encoding is high and visual quality deteriorates
Solution Approach 1:
The mesh is segmented into multiple patches, where each patch is processed independently for compression. This segmentation allows for more efficient encoding of displacement information within each patch, reducing the overall signaling cost while maintaining visual quality. The patch-based approach enables localized optimization of compression parameters.
Solution Approach 2:
Different compression strategies and precision levels are applied to different regions of the mesh based on local importance. Areas with higher visual significance receive more bits for displacement encoding, while less critical areas use coarser representation. This local quality adaptation reduces total signaling cost while preserving essential visual information.
2Manufacturing precision
If mesh subdivision is performed to increase fidelity, then visual quality improves, but data quantity and processing complexity increase
Solution Approach 1:
Mesh subdivision is performed preliminarily before compression, creating a refined mesh structure that captures fine geometric details. The displacement fields are then computed on this subdivided mesh, allowing high-fidelity representation to be achieved through efficient field encoding rather than storing complete high-resolution mesh data, thus reducing overall data quantity.
Solution Approach 2:
Instead of storing complete high-resolution mesh geometry in 3D space, the patent transitions to representing the mesh through displacement fields defined on a coarser reference mesh. This dimensional transformation converts the problem from storing numerous vertex positions to storing displacement vectors, significantly reducing data quantity while maintaining visual fidelity through the subdivision-improved displacement generation process.
Data Source
AI summary
A method of encoding includes receiving a polygon mesh that includes a plurality of vertices; subdividing the polygon mesh to generate a plurality sub-vertices; determining a first normal vector of a first vertex in the polygon mesh and a second normal vector of a second vertex in the polygon mesh, the first vertex and the second vertex defining an edge in the polygon mesh; moving a sub-vertex from the plurality of sub-vertices located on the edge using the first normal vector and the second normal vector; determining a displacement between the moved sub-vertex and a reference vertex from the plurality of vertices; and generating a bitstream including the displacement.