A framework for reconstruction based recognition of partially occluded repeated objects

In this paper we propose a reconstruction based recognition scheme for obje cts with repeated components, using a single image of such a configuration, in which one of the repeated components may be partially occluded. In our strategy we reconstruct each of the components with respect to the same fra me and use these to compute invariants.We propose a new mathematical framew ork for the projective reconstruction of affinely repeated objects. This us es the repetition explicitly and hence is able to handle substantial occlus ion of one of the components. We then apply this framework to the reconstru ction of a pair of repeated quadrics. The image information required for th e reconstruction are the outline conic of one of the quadrics and correspon dence between any four points which are images of points in general positio n on the quadric and its repetition. Projective invariants computed using t he reconstructed quadrics have been used for recognition. The recognition s trategy has been applied to images of monuments with multi-dome architectur e. Experiments have established the discriminatory ability of the invariant s.