3D Object Representation And Analysis With Application In Object Simulation And Retrieval

3D applications have been widely used in many fields ranging from industrial CAD/CAM design to medical simulation of organ deformation and to the entertainment industry in our daily life. More and more applications based on 3D scene come cross and give customers much better experiences. Although people have already got some sophisticated methods to reconstruct 3D objects from a set of 2D images, how to represent and analyze the objects in more effective ways still remains a problem, which is now becoming a hot research topic related to the areas of computer vision and machine learning, to which many researchers devote themselves.By taking advantage of the statistical and machine learning methods, the thesis provides effective solutions to applications of two categories:learning deforming patterns for object simulation and extracting descriptive features for objects retrieval. Highlight contributions are as follows1. For object simulation, a novel Statistical Deformable Model (Mr-SDM) is proposed to learn the deforming patterns related to the material properties from a set of man-made training samples which are generated through Finite Element Method. By combining SDM and FEM together, Mr-SDM speeds up the process of physical analysis through transforming the high computational cost of FEM to the process of offline training while at the same time, solves the small sample size problem of traditional SDM.2. A prediction method based on Mr-SDM is proposed to predict the facial changes in the orthognathic surgery. Such method is then extended to a generic deforming algorithm dealing with common objects with different shapes and material parameters.3. As for object retrieval, on the other hand, visual topics are proposed to extract the shape characters from the raw features by clustering the nearby vertices or views with similar behaviors together, which are further extended to multi-resolution topics to describe the object at different smooth levels to meet demands of various applications.4. 3D object retrieval based on topic representation is then presented, where the performances with different similarity measure have been investigated...