Structured Hough Voting for Vision-based Highway Border Detection.pdfVIP

Structured Hough Voting for Vision-based Highway Border Detection Zhiding Yu?, Wende Zhang?, B. V. K. Vijaya Kumar?, Dan Levi? ?Department of Electrical and Computer Engineering, Carnegie Mellon University ?Electrical and Controls Systems Research Lab, General Motors Company ?Advanced Technical Center - Israel, General Motors Company yzhiding@, {wende.zhang,dan.levi}@, kumar@ Abstract We propose a vision-based highway border detection algorithm using structured Hough voting. Our approach takes advantage of the geometric relationship between high- way road borders and highway lane markings. It uses a strategy where a number of trained road border and lane marking detectors are triggered, followed by Hough voting to generate corresponding detection of the border and lane marking. Since the initially triggered detectors usually re- sult in large number of positives, conventional frame-wise Hough voting is not able to always generate robust bor- der and lane marking results. Therefore, we formulate this problem as a joint detection-and-tracking problem under the structured Hough voting model, where tracking refers to exploiting inter-frame structural information to stabilize the detection results. Both qualitative and quantitative e- valuations show the superiority of the proposed structured Hough voting model over a number of baseline methods. 1. Introduction Detecting road borders has broad applications in future autonomous vehicles and intelligent transportation systems as an important component of scene understanding. It can provides cues about road structure that benefit motion plan- ning and cruise behavior control. In autonomous driving, the detection of road border is often done by GPS, high quality road map and some other active sensors such as Radar and Lidar. This, however, can sometimes be limit- ed by the accuracy of GPS positioning signal as well as the resolution of active sensors. A natural question is whether we can address the problem with computer vis