A Bayesian 3-D Search Engine Using Adaptive Views Clustering

Tarik Filali Ansary (1, 2) , Mohamed Daoudi (1, 2) , Jean-Philippe Vandeborre (1, 2)

1 LIFL - FOX MIIRE
2 Institut TELECOM/TELECOM Lille1

In this paper, we propose a method for three-dimensional (3-D)-model indexing based on two-dimensional (2-D) views, which we call adaptive views clustering (AVC). The goal of this method is to provide an "optimal" selection of 2-D views from a 3-D model, and a probabilistic Bayesian method for 3-D-model retrieval from these views. The characteristic view selection algorithm is based on an adaptive clustering algorithm and uses statistical model distribution scores to select the optimal number of views. Starting from the fact that all views do not have equal importance, we also introduce a novel Bayesian approach to improve the retrieval. Finally, we present our results and compare our method to some state-of-the-art 3-D retrieval descriptors on the Princeton 3-D Shape Benchmark database and a 3-D-CAD-models database supplied by the car manufacturer Renault.

Mots clés: Bayesian approach; clustering; 3-D indexing; 3-D retrieval; views

Domaines: Vision par ordinateur et reconnaissance de formes [cs.CV]

Tarik Filali Ansary, Mohamed Daoudi, Jean-Philippe Vandeborre. A Bayesian 3-D Search Engine Using Adaptive Views Clustering. IEEE Transactions on Multimedia, 2007, 9 (1), pp.78-88. ⟨10.1109/TMM.2006.886359⟩. ⟨hal-00666134⟩