Dominant points detection for 3D vision

Abstract

An algorithm for detecting dominant points (corners and trees) in images of 3D objects is presented. The technique chooses ideal points from the basic definitions of both tree and corner points. The algorithm has two subiterations, in the first one the initial candidate dominant points are determined, and in the second subiteration the candidate dominant points are tested. Experiments were performed to show that the proposed algorithm is reliable for 3D object recognition by meaning of representing independent view point images of the same object by a set of dominant points.