LINE CORRESPONDENCES FROM COOPERATING SPATIAL AND TEMPORAL GROUPING PROCESSES FOR A SEQUENCE OF IMAGES

This paper addresses the problem of matching line segments in two imag es from a sequence of images. We propose a new algorithm that employs relaxation labeling to integrate the perceptual grouping and feature m atching processes. We consider feature matching between two views as a ''temporal grouping'' process, in addition to the traditional spatial groups established by perceptual grouping in a single image. In the r elaxation procedure, we compare a line in the first image with multipl e matching candidates in the second image. Correspondence ambiguities are resolved by iteratively propagating constraint information from th e neighbors of a line, which are defined by the line's perceptual grou ps. The applications of our work include moving object detection, imag e registration, mobile robot navigation, and 3D structure from motion. We discuss the new algorithms and present performance analysis and ex perimental results with real image sequences. (C) 1997 Academic Press.