3D Object Retrieval With Multi-feature Collaboration And Bipartite Graph Matching

Recently, extensive research efforts have been dedicated to view-based methods for 3-D object retrieval due to the highly discriminative property of multiviews for 3-D object representation. However, most of state-of-the-art approaches highly depend on one descriptor to express views of 3-D objects. In order to break up the limitation of a single descriptor,we propose a novel 3D object retrieval with features collaboration and bipartite graph matching strategies which explored the essential characters of 3D object in a view-based retrieval framework. By extracting complement descriptors from both the contour and the interior region of 3D object effectively. Specifically, Zernike Moments and Bo VW/SIFT collaborate in the region shape description; a boundary chain code based Fourier descriptor is explored to model the contour context of 3D object. With the bipartite graph matching and feature concatenation, as well as the ad option of user feedback strategy, significant performance improvement is achieved in the 3D object retrieval task.Aiming at utilizing the collaborated features for multiple views and conducting multi view matching and estimating the relevance among diffe rent 3D objects, a novel 3D object retrieval has been proposed with features collaboration and bipartite graph matching strategies, and the main contributions of this paper are summarized as follows:Inspired by the essential characters of 3D object, a view-based retrievalframework with multi-feature collaboration and bipartite graph matching isproposed, which extracts complement descriptors from both the contourand the interior region of 3D object effectively.A Greedy Search(GS) algorithm is proposed to calculate the similarity ofquery object and object and three bipartite graphs are employed to obtainthe optimal match of each bipartite graph pair.A feedback based re-ranking strategy is adopted to improve theperformance of 3D object retrieval.Our proposed method participated the SHREC’15 challenge and achieved the state-of-the-art performance.