Mechanical Mesh Segmentation and Global 3D Shape Extraction

摘要: This paper presents a method for segmenting noisy 2-manifold meshes based on decomposition into local shape primitives maximizing global coherence. technique works by partitioning the input mesh regions which can be approximated simple geometrical primitive such as plane, sphere or cylinder. The proposed approach is entirely error-driven, convergence-proven, and does not need to specify number of segments. The guided robust extractions RANSAC sampling graphical model regularizes segmented regions. final minimum energy associated with this model. Obtained segmentations mechanical outperform other approaches in terms region contour correctness consistency object decomposition. Applications work are reverse engineering, structure analysis, feature enhancement, noise removal, compression, piecewise approximation geometry, remeshing.