This thesis describes an online search engine for 3D models, focusing on query interfaces and their corresponding model/query representations and matching methods. A large number of 3D models has already been created, many of which are freely available on the web. Because of the time and effort involved in creating a high-quality 3D model, considerable resources could be saved if these models could be reused. However, finding the model you need is not easy, since most online models are scattered across the web, on repository sites, project sites, and personal homepages. To make these models more accessible, we have developed a prototype 3D model search engine. This project serves as a test bed for new methods in web crawling, query interfaces, and matching of 3D models. This thesis focuses on query interfaces and their accompanying matching methods. We investigated query interfaces based on text keywords, 3D shape, 2D shape, and some combinations. By testing matching methods that use text, 3D shape, and 2D shape, we found that the 3D shape matching method outperforms our text matching method for our application domain, due to the insufficient text annotation of 3D models on the web. Furthermore, classification performance was improved by combining the 3D shape and text-based matching methods. The results of a user study also suggest that text can combine with shape to make queries more effective. We compared shape matching methods based on matching multiple 2D projections of a 3D model and found that from a set of side, corner and edge projections, the combination of side and corner projections produced the best results. However, these results were still worse than those of the 3D shape matching method. We present a 2D structural interface and accompanying matching method that produces better classifications than image matching for certain types of objects. Our prototype search engine has been used extensively across the world (in 18 months, almost 300,000 queries have been processed from more than 100 countries) and has proven to be useful (models have been downloaded over 50,000 times, and almost 30% of all visitors per day are returning users).

@phdthesis{:2004:A3M,
   author = "Patrick Min",
   title = "A {3D} Model Search Engine",
   school = "Princeton University",
   year = "2004",
   month = jan
}