A tracker for broken and closely spaced lines
Naoki Chiba, Takeo Kanade · Systems and Computers in Japan · 1999
We propose an automatic line tracking method that is robust over a real image sequence. In a real image sequence, it is difficult to track broken and closely spaced lines correctly. We solve this problem with the following three techniques. First, by using a hierarchical optical-flow estimation technique, we can obtain a good prediction of line segment motion in a consecutive frame. Second, we propose a matching similarity function to perform multiple collinear-line segment matching, instead of merely one-to-one matching. Third, we propose to use edge directions to discriminate closely spaced lines. Experiments using noisy and complicated real image sequences taken with a hand- held camcorder confirm the robustness of our method in difficult tasks. © 1999 Scripta Technica, Syst Comp Jpn, 30(7): 18–26, 1999