SEMILOCAL PROJECTIVE INVARIANTS FOR THE RECOGNITION OF SMOOTH PLANE-CURVES

Recently, several methods have been proposed for describing plane, non -algebraic curves in a projectively invariant fashion. These curve rep resentations are invariant under changes in viewpoint and therefore id eally suited for recognition. We report the results of a study where t he strengths and weaknesses of a number of semi-local methods are comp ared on the basis of the same images and edge data. All the methods de fine a distinguished or canonical projective frame for the curve segme nt which is used for projective normalisation. In this canonical frame the curve has a viewpoint invariant signature. Measurements on the si gnature are invariants. All the methods presented are designed to work on real images where extracted data will not be ideal, and parts of c urves will be missing because of poor contrast or occlusion. We compar e the stability and discrimination of the signatures and invariants ov er a number of example curves and viewpoints. The paper concludes with a discussion of how the various methods can be integrated within a re cognition system.