MODEL-BASED OBJECT RECOGNITION - THE ROLE OF AFFINE INVARIANTS

This paper proposes an efficient method to recognize rigid flat object s from its intensity images which are assumed to be arbitrarily positi oned in space. The task of the recognition method is to find instances of known object models in affine images. Affine invariant shape descr iptors of rigid flat objects are generated which are invariant to chan ge in the point of view. In the proposed paradigm, the objects are des cribed by sets of local and global features. Since we are also concern ed with the recognition of partially occluded objects, the local featu res are given importance for obtaining descriptions of objects. The gl obal features are useful for finding the exact match and are used for verification. The local features can be points, line segments, curve s egments, etc. We restrict ourselves to points, which are referred to a s interest points. The point set of the various model objects are matc hed simultaneously against the point set of the composite overlapping scene using a small number of corresponding points. Seven discrete mom ents are used here as global features which are also invariant under t he affine transformation. Experiments show good performance and togeth er with inherent parallelism of the recognition method makes the metho d a promising one.