Line Correspondences from Cooperating Spatial and Temporal Grouping Processes for a Sequence of Images

hypotheses may grow exponentially. To address this problem, it is clear that more a priori information is needed. One This paper addresses the problem of matching line segments in two images from a sequence of images. We propose a new way to incorporate such information is to make implicit algorithm that employs relaxation labeling to integrate the perassumptions about objects in motion, such as the local ceptual grouping and feature matching processes. We consider smoothness of motion fields, which serves as the basis for feature matching between two views as a ‘‘temporal grouping’’ most feature matching algorithms. process, in addition to the traditional spatial groups established In this paper we propose a new approach for image by perceptual grouping in a single image. In the relaxation matching based on generalized perceptual grouping and procedure, we compare a line in the first image with multiple relaxation labeling. We choose to use line features because matching candidates in the second image. Correspondence ammanmade objects usually have line features and lines can biguities are resolved by iteratively propagating constraint inbe accurately extracted by well-known algorithms. Our formation from the neighbors of a line, which are defined by the system is thus suitable for structured environments such line’s perceptual groups. The applications of our work include moving object detection, image registration, mobile robot navias indoor or urban scenes. The applications of our work gation, and 3D structure from motion. We discuss the new include moving object detection, map image registration, algorithms and present performance analysis and experimental mobile robot navigation, and computation of 3D structures results with real image sequences.  1997 Academic Press from motion [3]. Our proposed system considers matching as a temporal grouping process, and we adopt the statistic-based ap