상세정보
학술대회
Robust Lane Detection for Video-Based Navigation Systems
- 발행일
- 200710
- 출처
- International Conference on Tools with Artificial Intelligence (ICTAI) 2007, pp.535-538
- 협약과제
- 07MD1900, 텔레매틱스용 실감컨텐츠 구축/관리 기술 개발, 조성익
- 초록
- Lane detection from a live video captured in a moving vehicle is an important issue for autonomous vehicles and video-based navigation systems. In this paper, we present a novel idea for robust lane detection and lane color recognition. More specifically, a framework for robust lane detection is presented. Then, a novel idea to reduce illumination effects is presented. Lastly, SVM approach is presented to recognize lane color robustly for various lighting conditions including shadow, backlight, sunset, and so on. By combining information from navigation database, it is possible to decide if we are in the leftmost, middle, or the rightmost lane, which allows us to provide more realistic navigation information to drivers. Simulation results are provided to show the robustness of the proposed idea. © 2007 IEEE.
- KSP 제안 키워드
- Autonomous vehicle, Color Recognition, Combining information, Lane Detection, Lighting conditions, Moving Vehicle, Navigation database, Navigation information, Rightmost lane, live video, navigation system