3D-PSSIM: Projective Structural Similarity for 3D Mesh Quality Assessment Robust to Topological Irregularities.

Area of Science:

• Computer Graphics
• Image Processing
• Perceptual Computing

Background:

• Assessing 3D mesh quality is challenging due to irregular topology.
• Existing algorithms often fail to correlate with human subjective opinions.
• Defining robust mesh quality features is difficult.

Purpose of the Study:

• To develop a novel 3D mesh quality assessment algorithm.
• To create a method robust to topological differences and perceptual distortions.
• To achieve high correlation with human subjective judgments.

Main Methods:

• Proposed a 3D projective structural similarity index (3D-PSSIM).
• Utilized multi-view and multi-layer projections to represent mesh geometry and texture.
• Incorporated visual sensitivity weights for surface curvature perception.
• Aggregated quality-aware features from multiple projective spaces onto the mesh domain.

Main Results:

• 3D-PSSIM demonstrates high correlation with human subjective quality assessments.
• The method is robust to noise and significant topological variations.
• Outperformed existing mesh quality assessment models in experiments.

Conclusions:

• 3D-PSSIM offers a reliable and perceptually aligned approach for 3D mesh quality assessment.
• The projective method effectively handles topological irregularities and occlusion.
• This technique advances the evaluation of 3D models in computer graphics.