Robust Mesh Segmentation Using Feature-Aware Region Fusion.

Area of Science:

• Computer Graphics
• Geometric Modeling
• Computational Geometry

Background:

• 3D mesh segmentation is crucial for various applications, including computer graphics, virtual reality, and engineering.
• Existing algorithms often struggle with complex geometries or lack robustness and efficiency.

Purpose of the Study:

• To introduce a novel, simple, yet powerful segmentation algorithm for 3D meshes.
• To enhance the robustness and efficiency of 3D mesh segmentation for complex models.

Main Methods:

• The algorithm employs a two-stage approach: over-segmentation via adaptive space partition and iterative region fusion.
• Region fusion utilizes novel intra-region and inter-region difference metrics, incorporating various shape features.
• The feature-aware fusion process ensures adaptability to complex mesh structures.

Main Results:

• The proposed algorithm demonstrates high efficiency in the over-segmentation stage.
• The feature-aware region fusion effectively handles complex 3D meshes, achieving robust segmentation.
• Extensive qualitative and quantitative experiments confirm the algorithm's advantages over existing methods.

Conclusions:

• The developed segmentation algorithm offers a robust and efficient solution for 3D mesh processing.
• Its feature-aware approach makes it suitable for segmenting intricate and complex 3D models.
• The algorithm's performance is well-supported by comprehensive experimental validation.