Junction detection based on line segments

We present a novel method for junction detection. A junction is defined as the point where several lines intersect. Given the line segments in the image, our junction detector consists of three steps. First, potential junctions are located from a small neighborhood around the intersection of each pair of lines. Second, our detector searches the branches connecting to each potential junction in a circular neighborhood according to its scale. Finally, the actual junctions are selected using a two-step method according to their connecting branches and their distance. We test our algorithm on a variety of images. The experiment results demonstrate that the proposed algorithm can robustly detect junctions in different scenes.