A Method of Similarity Retrieval for Three-Dimensional Shape by Using Spatial Partitioning Model

Abstract

Demands for similarity retrieval of three-dimensional model have increased because of the spread of digital archive technology. This paper proposes a new method of similarity estimation by using spatial partitioning model and mathematical morphology. First, our method extracts an amount of global feature of the shape by using spatial partitioning model and local feature of the shape by mathematical morphology. Secondly, the method calculates similarity between the amount of the feature and those which are preserved in advance. Some experimental results show that our method is more effective than the conventional method.