The CAD model retrieval based on shape similarity is the research focus of computer graphics and computer-aided design field. To obtain higher retrieval precision and efficiency, a new CAD model retrieval method based on shape similarity is proposed. The method is divided into two steps. Firstly, breadth-first-search-based spanning tree algorithm is applied to obtain an initial boundary matching between retrieval object and retrieval condition. Secondly, the topology adjacency approximation algorithm is put forward to find the optimal boundary matching based on the initial boundary matching by the cycle and approximation process and the matching result is used to calculate the shape similarity between retrieval object and retrieval condition. In order to calculate the similarity between various types of boundary faces, a new similarity calculation method is present. Finally, a CAD model retrieval system on the platform of UG is developed based on the proposed method. Experimental results show that the proposed method is feasible and effective.
展开▼
