Spherical Hybrid Curvature Images of 3D Shapes and Its Applications

Abstract

A new representation of 3D shapes is proposed using spherical hybrid curvature images. Signals like normals and curvatures are stored in spherical hybrid curvature images using spherical parameterization and partition of a sphere. The generation depends on several geometric signals, not one single geometric signal. So its colors reflect the properties of 3D shapes more clearly. With simple arrays similar to 2D image, the spherical hybrid curvature image can be post-processed easily using sophisticated image processing algorithms. Applications include segmentation, editing and model retrieval etc. Experimental results show that our method is better than other geometry image algorithms. The algorithm is simple and practical especially for models in engineering.