Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Textured mesh Quality Assessment using Geometry and Color Field Similarity

Kaifa Yang, Qi Yang, Zhu Li, Yiling Xu

TL;DR

This work introduces FMQM, a full-reference TMQA metric designed for textured meshes by using field-based representations that couple geometry and texture via SDF and a novel color field called NCF. FMQM explicitly models both geometric and color-field differences through four perceptually motivated features—geometry similarity, geometry gradient similarity, space color distribution similarity, and space color gradient similarity—and aggregates them with a fourth-root pooling to obtain the final score. The method achieves state-of-the-art correlations across three benchmark datasets (TSMD, SJTU-TMQA, YANA) with robust performance under diverse distortions and lower computational cost due to offline patching and near-surface sampling. The results demonstrate FMQM’s practical applicability for real-time TMQA in 3D graphics and visualization, while the NCF-based field representation offers a flexible framework for future field-based metric enhancements. $FMQM = (geoSSIM \cdot geoGraSSIM \cdot colorSSIM \cdot colorGraSSIM)^{1/4}$ serves as the central aggregation formula linking geometry and color cues to perceptual quality.

Abstract

Textured mesh quality assessment (TMQA) is critical for various 3D mesh applications. However, existing TMQA methods often struggle to provide accurate and robust evaluations. Motivated by the effectiveness of fields in representing both 3D geometry and color information, we propose a novel point-based TMQA method called field mesh quality metric (FMQM). FMQM utilizes signed distance fields and a newly proposed color field named nearest surface point color field to realize effective mesh feature description. Four features related to visual perception are extracted from the geometry and color fields: geometry similarity, geometry gradient similarity, space color distribution similarity, and space color gradient similarity. Experimental results on three benchmark datasets demonstrate that FMQM outperforms state-of-the-art (SOTA) TMQA metrics. Furthermore, FMQM exhibits low computational complexity, making it a practical and efficient solution for real-world applications in 3D graphics and visualization. Our code is publicly available at: https://github.com/yyyykf/FMQM.

Textured mesh Quality Assessment using Geometry and Color Field Similarity

TL;DR

This work introduces FMQM, a full-reference TMQA metric designed for textured meshes by using field-based representations that couple geometry and texture via SDF and a novel color field called NCF. FMQM explicitly models both geometric and color-field differences through four perceptually motivated features—geometry similarity, geometry gradient similarity, space color distribution similarity, and space color gradient similarity—and aggregates them with a fourth-root pooling to obtain the final score. The method achieves state-of-the-art correlations across three benchmark datasets (TSMD, SJTU-TMQA, YANA) with robust performance under diverse distortions and lower computational cost due to offline patching and near-surface sampling. The results demonstrate FMQM’s practical applicability for real-time TMQA in 3D graphics and visualization, while the NCF-based field representation offers a flexible framework for future field-based metric enhancements. serves as the central aggregation formula linking geometry and color cues to perceptual quality.

Abstract

Textured mesh quality assessment (TMQA) is critical for various 3D mesh applications. However, existing TMQA methods often struggle to provide accurate and robust evaluations. Motivated by the effectiveness of fields in representing both 3D geometry and color information, we propose a novel point-based TMQA method called field mesh quality metric (FMQM). FMQM utilizes signed distance fields and a newly proposed color field named nearest surface point color field to realize effective mesh feature description. Four features related to visual perception are extracted from the geometry and color fields: geometry similarity, geometry gradient similarity, space color distribution similarity, and space color gradient similarity. Experimental results on three benchmark datasets demonstrate that FMQM outperforms state-of-the-art (SOTA) TMQA metrics. Furthermore, FMQM exhibits low computational complexity, making it a practical and efficient solution for real-world applications in 3D graphics and visualization. Our code is publicly available at: https://github.com/yyyykf/FMQM.
Paper Structure (41 sections, 34 equations, 10 figures, 5 tables)

This paper contains 41 sections, 34 equations, 10 figures, 5 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Projection-based Methods
  4. Model-based Methods
  5. Point-based Methods
  6. Summary
  7. Preliminary
  8. TMQA Problem Formulation
  9. Field Description
  10. NCF: A New Color Field Description for TMQA
  11. Motivation and Definition
  12. NCF Property
  13. NCF Feature Extraction Unit
  14. Method
  15. Framework
  16. ...and 26 more sections

Figures (10)

  • Figure 1: Projection Views of drawingRoom from TSMDtsmd and bennetTomb from SJTU-TMQAsjtumqa
  • Figure 2: Masking Effect Caused by Mesh Texture
  • Figure 3: Definition and Properties of NCF
  • Figure 4: Framework of FMQM (2D illustration; 3D visualization provided in supplementary material)
  • Figure 5: Per-model correlation analysis
  • ...and 5 more figures