Extensive research efforts have been dedicated to 3D model retrieval inrecent decades. Recently, view-based methods have attracted much researchattention due to the high discriminative property of multi-views for 3D objectrepresentation. In this report, we summarize the view-based 3D model methodsand provide the further research trends. This paper focuses on the scheme formatching between multiple views of 3D models and the application ofbag-of-visual-words method in 3D model retrieval. For matching between multipleviews, the many-to-many matching, probabilistic matching and semisupervisedlearning methods are introduced. For bag-of-visual-words application in 3Dmodel retrieval, we first briefly review the bag-of-visual-words works onmultimedia and computer vision tasks, where the visual dictionary has beendetailed introduced. Then a series of 3D model retrieval methods by usingbag-of-visual-words description are surveyed in this paper. At last, wesummarize the further research content in view-based 3D model retrieval.
展开▼
