PARTIAL SURFACE AND VOLUME MATCHING IN 3 DIMENSIONS

In this paper we present a new technique for partial surface and volum e matching of images in three dimensions. In this problem we are given two objects in 3-space, each represented as a set of points, and the goal is to find a rigid motion of one object which makes a sufficientl y large portion of its boundary lying sufficiently close to a correspo nding portion of the boundary of the second object. This is an importa nt problem in pattern recognition and in computer vision, with many in dustrial, medical, and chemical applications. Our method treats separa tely the rotation and the translation components of the Euclidean moti on that we seek. The algorithm steps through a sequence of rotations, in a steepest-descent style, and uses a novel technique for scoring th e match for any fixed rotation. Experimental results on various exampl es, involving data from industrial applications, medical imaging, and molecular biology, are presented, and show the accurate and robust per formance of our algorithm.