3d Model Retrieval In Several Feature Extraction Method

With the development of 3D models acquisition, computer graphics, and computer network technology, 3D models are more and more widely used in many fields. Taking into account that building the 3D models is a very complex work, it is very important to reuse 3D models. How to retrieve 3D models needed by customers quickly from the huge database is becoming more and more necessary.Generally speaking, a complete 3D model retrieval system typically includes feature extraction, similarity matching, index structure, and query interface. And feature extraction is the most important factor for the 3D model retrieval. So 3D model feature extraction is the key technology in 3D model retrieval, and it is also the main point of this paper. The main job of this paper is to design and implement three new methods of feature extraction through researching the technology of 3D model feature extraction.1. 2D projective point sets. It is different from the traditional methods, which compares the 3D shapes based on 2D contour map, and compares the 3D shapes by measuring the statistical characteristic of 2D projective point sets. In fact, this method uses the technology of 2D image retrieval, and it has low complexity.2. Achieve and improve on the thearithmetic of the transform of sphere congruity , and with the traditional methods Compared , it has less computation.3. The feature extraction of skeleton. Design and implement a better algorithm of the feature extraction of skeleton: use the method is distance's change, And get the information combining to the arithmetic of connectivity judgement, achieve the feature extraction of Skeleton.The paper have developed a 3D model retrieval system and done a lot of experiments based on all the three arithmetic.the experimental results show that the proposed feature extraction algorithms are effective and efficient. They can distinguish different 3D model from each other.