3D object model retrieval

Abstract

Pattern recognition is a fundamental problem. With large number of 3D models available on the web, and the application of 3D object models in computer graphics, computer vision and molecular biology, an effective way to search the 3D model are important. Recent year, different 3D object model search engine is available. They are based on different algorithms. In this project, I will use the features generated by different algorithms to do the similarity search. Neural network and different classifier were used to classify the objects. I also propose a new method to do normalization and features generation which provide an alternative to the existing methods.