Chroma Displacement Field Upsampling for Dynamic Mesh Compression

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing mesh compression standards struggle to efficiently handle dynamic meshes with time-varying connectivity and attribute maps, particularly in real-time applications like augmented and virtual reality, lacking support for lossy and lossless compression.

Innovation Solution

The proposed method involves generating a down-sampled displacement field for a chroma component, reconstructing and up-sampling it using additional displacement vectors, including zero vectors and interpolated vectors based on distance and weight, to enhance mesh decoding and encoding efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If down-sampled displacement field is used for chroma component, then compression efficiency is improved, but decoding quality deteriorates

Engineering Contradiction:
Improvecompression efficiencyVSAvoiddecoding quality
Core Design Contradiction:
ProductivityVSManufacturing precision

Solution Approach 1:

The displacement field is segmented into luma and chroma components, with the chroma component being down-sampled to reduce data量. The chroma displacement field is processed separately at a lower resolution, allowing compression efficiency improvement while maintaining overall quality through selective processing.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

Different quality levels are applied to different components: luma displacement field maintains full resolution while chroma displacement field uses down-sampled resolution. This local quality differentiation improves compression efficiency by reducing the chroma data量 while preserving critical luma information that has greater visual impact.

Inventive Principle:
Principle #3Local quality

Solution Approach 3:

The chroma displacement field is down-sampled in advance before encoding, and the decoder is pre-configured with knowledge of the down-sampling format. This preliminary action allows the decoder to efficiently reconstruct the displacement field without requiring complex real-time processing, thus improving compression efficiency while maintaining acceptable quality.

Inventive Principle:
Principle #10Preliminary action

2Manufacturing precision

If additional displacement vectors are added through interpolation, then decoding quality is improved, but device complexity increases

Engineering Contradiction:
Improvedecoding qualityVSAvoidprocessing complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The decoder copies the down-sampled chroma displacement field and uses it directly for prediction without requiring complex interpolation algorithms. By copying and reusing the down-sampled data appropriately, the system maintains decoding quality while avoiding the computational complexity of generating additional interpolated vectors.

Inventive Principle:
Principle #26Copying

Solution Approach 2:

The system changes the resolution parameter of the chroma displacement field from full resolution to down-sampled resolution. This parameter change reduces the number of vectors that need to be processed and stored, thereby improving decoding quality through efficient use of available data while reducing processing complexity compared to handling full-resolution chroma data.

Inventive Principle:
Principle #35Parameter changes

3Adaptability or versatility

If time-varying connectivity and attribute maps are supported, then adaptability is improved, but device complexity increases

Engineering Contradiction:
Improvesupport for dynamic meshesVSAvoidprocessing complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The mesh compression system is designed to handle dynamic meshes where connectivity and attribute maps vary over time. By using video coding techniques with temporal prediction, the system adapts to time-varying properties by predicting current frame characteristics from previous frames, improving adaptability for real-time applications while managing processing complexity through efficient temporal reuse of data.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The displacement field mechanism is designed to be universal, serving multiple functions: it handles both static and dynamic meshes, supports various sampling formats, and works with different chroma subsampling configurations. This multi-functionality improves adaptability for diverse applications including AR/VR while consolidating processing requirements rather than adding separate specialized systems.

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

Data Source

PatentUS12425648B2Upsampling of displacement field in mesh compression
Publication Date: 2025.09.23 TENCENT AMERICA LLC
  • US12425648B2 patent drawing
  • US12425648B2 patent drawing
  • US12425648B2 patent drawing

AI summary

Aspects of the disclosure includes methods and apparatuses for mesh processing and a method of processing mesh data. The apparatus for mesh processing includes processing circuitry configured to receive coded information indicating that a displacement field of a chroma component of a current frame is down-sampled according to a sampling format. The down-sampled displacement field of the chroma component includes a first array of displacement vectors. The processing circuitry is configured to reconstruct the first array of displacement vectors of the down-sampled displacement field and up-sample the down-sampled displacement field according to the sampling format to generate an up-sampled displacement field of the chroma component that includes a second array of displacement vectors. The second array of displacement vectors includes the first array of displacement vectors and additional displacement vectors.