3D Model Retrieval Method Based on Single View

In order to solve the view redundant problem in existing research work, we propose an approach based on a single view for 3D model shape feature measuring, and the similarity between models is calculated through dynamic programming algorithm. This method implements single view feature description and shape matching for 3D model. The algorithm is composed of three steps: Firstly, the pose of a 3D model is adjusted and the main view of the 3D model that best describes the profile features of the 3D model can be acquired by rendering. Then, through contour sampling of the rendered view, the model shape feature is described through extracting inner distance and inner angle. Finally, the similarity between different 3D models is calculated through dynamic programming algorithm. The experimental results show that the retrieval precision of the proposed method based on a single view is better than some traditional retrieval methods using 3D shape feature descriptor.