3D model search engine

Abstract

Large number of 3D models available on the Internet. And the application of 3D model in different fields such as computer game industry, manufacturing, medical application become very common. An effective 3D model search engine is required for searching in huge amount of 3D models. Some 3D model search engines have been developed with different searching algorithm. In this project, shape descriptor of 3D models parts were extracted by calculating density of point clouds in each equivalence class in the order of 90 degrees, with multiple bounding boxes applied on each 3D model parts and its percentage of total area size. 3D model matching and retrieval was done by the query of partial shape of 3D model. The relation of area size difference, Euclidian Distance and precision of searching result was investigated.