Based On The Content Of 3d Model Retrieval System Design And Realization

With the development of three-dimensional modeling techniques and computer hardware/software technologies, the application of 3D models is more and more widely and the number of 3D models is growing larger and larger. Makeing full use of the existing 3D data resources, can greatly reduce the workload of the design of the new models. At present, the keywords based 3D model retrieval method has been unable to meet the requirements of the design and practical applications. The content-based 3D model retrieval technology has broken through the traditional limitations of keyword search by directly making use of the characteristics of 3D model which creates a indexing for retrieval. Its main advantages include objectivity, information retrieval and high-speed retrieval.The key point of Content-based 3D model retrieval is how to quantify the 3D model's features based on which defines the 3D model's similarity. As the 3D model after similar transformations is still regarded as the same one, it's necessary to extract translation, rotation and scale invariant geometric features. Therefore, the definition of the 3D model's shape descriptor is expected to satisfy rotation invariance.Shape distribution is one of the statistical-based 3D model feature extraction methods. It is easy to compute, and robust against 3D model simplification and noise. Especially, it meets the translation, rotation and scale invariance. In this thesis, a content-based 3D model retrieval system is designed using the shape distribution to extract shape features. The system is able to search accurate results according user requirements. The studying of content-based 3D model retrieval system is of great value both in theoretical significance and practical application.