3D Object Recognition Using Hyper-Graphs and Ranked Local Invariant Features
Local invariant feature-based methods such as SIFT have been proven highly effective for object recognition. However, they have made either relatively little use or too complex use of geometric constraints and are confounded when the detected features are superabundant. Here we make two contributions aimed at overcoming these problems. First, we rank the SIFT points (R-SIFT) using visual saliency. Second, we use the reduced set of R-SIFT features to construct a class specific hyper graph (CSHG) which comprehensively utilizes local SIFT and global geometric constraints. Moreover, it efficiently captures multiple object appearance instances. We show how the CSHG can be learned from example images for objects of a particular class. Experiments reveal that the method gives excellent recognition performance, with a low false-positive rate.
- 4.Chung, F.: Spectral Graph Theory. American Mathematical Society (1997)Google Scholar
- 7.Li, F.F., Perona, P.: A Bayesian hierarchical model for learning natural scene categories. CVPR 2, 524–531 (2005)Google Scholar
- 10.Yan, K., Sukthankar, R.: PCA-SIFT: A more distinctive representation for local image descriptors. CVPR (2), 506–513 (2004)Google Scholar
- 12.Sivic, J., Zisserman, A.: VideoGoogle: A text retrieval approach to object matching in videos. ICCV 2, 1470–1477 (2003)Google Scholar
- 13.Sivic, J., Russell, B.C., Efros, A.A., Zisserman, A., Freeman, W.T.: Discovering objects and their location in images. ICCV 1, 370–378 (2005)Google Scholar