Abstract
Lots of rear end collisions due to driver inattention have been identified as a major automotive safety issue. A short advance warning can reduce the number and severity of the rear end collisions. This paper describes a Forward Collision Warning (FCW) system based on monocular vision, and presents a new vehicle detection method: appearance-based hypothesis generation, template tracking-based hypothesis verification which can remove false positive detections and automatic image matting for detection refinement. The FCW system uses time to collision (TTC) to trigger the warning.In order to compute time to collision (TTC), firstly, haar and adaboost algorithm is utilized to detect the vehicle; Secondly, we use simplified Lucas-Kanade algorithm and virtual edge to remove false positive detection and use automatic image matting to do detection refinement; Thirdly, hierarchical tracking system is introduced for vehicle tracking; Camera calibration is utilized to get the headway distance and TTC at last. The use of a single low cost camera results in an affordable system which is simple to install. The FCW system has been tested in outdoor environment, showing robust and accurate performance.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.