Along with the number of 3D models used on the web or stored in databases increasing rapidly, it becomes necessary for the users to retrieve a similar 3D model from huge amount of 3D models. Since the shape-based 3D model retrieval does not include semantics concept , so the result of retrieval can no longer meet the needs of the users. For this phenomenon, we propose a new 3D model retrieval method which combines the semantics with the shape features. K-means algorithm is employed to cluster the shape features to the semantic group. Spatial relationships are used to disambiguate among models with similar appearance. The Princeton Shape Benchmark Database is used to interpret the experimental results, and it demonstrates the feasibility of the method.
国家科技期刊平台
NSTL
维普
万方数据
