Dynamic 3D Mesh Compression with Material Indices

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing video coding and decoding standards struggle to efficiently compress dynamic 3D meshes with time-varying attribute maps and connectivity information, particularly under real-time constraints.

Innovation Solution

A method for compressing mesh sequences with multiple texture maps per frame, involving the determination of material indices for each triangle face and encoding these indices to efficiently represent and transmit dynamic mesh data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of energy

If existing mesh compression standards (IC, MESHGRID, FAMC) are used to compress dynamic meshes with constant connectivity, then compression efficiency is improved, but the standards cannot handle time-varying attribute maps and connectivity information

Engineering Contradiction:
Improvecompression efficiencyVSAvoidsupport for time-varying attribute maps and connectivity
Core Design Contradiction:
Loss of energyVSAdaptability or versatility

Solution Approach 1:

The patent extends existing compression standards from static constant-connectivity meshes to dynamic meshes with time-varying connectivity and attribute maps. The method dynamically determines material indices for each triangle face based on current mesh state and encodes these indices, enabling the system to adapt to changing mesh topologies and material assignments while maintaining compression efficiency.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent introduces material index parameters to represent time-varying attribute maps and connectivity information. By converting complex material and connectivity data into indexed representations, the system can efficiently encode and decode dynamic mesh variations without requiring complete retransmission of all mesh data.

Inventive Principle:
Principle #35Parameter changes

2Speed

If volumetric acquisition techniques are used to generate dynamic meshes with constant connectivity, then real-time performance is improved, but generating such meshes remains challenging under real-time constraints

Engineering Contradiction:
Improvereal-time generation speedVSAvoidcomplexity of volumetric acquisition processing
Core Design Contradiction:
SpeedVSDevice complexity

Solution Approach 1:

The patent extracts and encodes only the essential material index information for each triangle face rather than transmitting complete material and connectivity data. This selective encoding approach reduces the computational burden during transmission and decoding, making real-time processing more feasible even when working with complex volumetric acquisition data.

Inventive Principle:
Principle #2Taking out (Extraction)

3Quantity of substance

If glTF with Draco compression is used to reduce the size of 3D assets, then storage efficiency is improved, but the standard does not support dynamic meshes with time-varying connectivity and attribute maps

Engineering Contradiction:
Improvestorage efficiencyVSAvoidsupport for dynamic meshes with time-varying connectivity
Core Design Contradiction:
Quantity of substanceVSAdaptability or versatility

Solution Approach 1:

The patent creates a universal compression method that can handle both static and dynamic meshes, as well as meshes with constant and time-varying connectivity. By incorporating material index determination and encoding for each triangle face, the system provides multi-functional support for various mesh types while maintaining efficient compression ratios suitable for modern 3D applications.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentUS20250069275A1On compression of a mesh with multiple texture maps
Publication Date: 2025.02.27 TENCENT AMERICA LLC
  • US20250069275A1 patent drawing
  • US20250069275A1 patent drawing
  • US20250069275A1 patent drawing

AI summary

A method and apparatus comprising computer code for mesh coding configured to cause a processor or processors to receive an input dynamic mesh representing a volumetric data of at least one three-dimensional (3D) visual content, wherein the input dynamic mesh comprises a plurality of mesh frames; determine that a mesh frame among the plurality of mesh frames comprises a plurality of texture maps in response to a mesh file associated with the mesh frame indicating that at least two different materials are applied in the mesh frame; determine a material index associated with each triangle face in the mesh frame, wherein a respective material index indicates a texture to be applied to a respective triangle face; and encode material indices associated with the mesh frame of the input dynamic mesh.