STRUCTURAL MATCHING IN COMPUTER VISION USING PROBABILISTIC RELAXATION

In this paper, we develop the theory of probabilistic relaxation for m atching features extracted from 2D images, derive as limiting cases th e various heuristic formulae used by researchers in matching problems, and state the conditions under which they apply. We successfully appl y our theory to the problem of matching and recognizing aerial road ne twork images based on road network models and to the problem of edge m atching in a stereo pair. For this purpose, each line network is repre sented by an attributed relational graph where each node is a straight line segment characterized by certain attributes and related with eve ry other node via a set of binary relations.