3D shape retrieval using viewpoint information-theoretic measures
Xavier Bonaventura, Jianwei Guo, Weiliang Meng, Miquel Feixas, Xiaopeng Zhang and Mateu Sbert
Article first published online: 18 DEC 2013 | DOI: 10.1002/cav.1566
We present an information-theoretic framework to compute the shape similarity between 3D polygonal models. Three different methods are introduced to evaluate the shape similarity, and they are respectively based on the L2 distance between information spheres, the earth mover's distance between information histograms and the absolute difference between the mutual information of each model. The performance of these techniques is tested using the Princeton Shape Benchmark database.