This paper proposes an efficient approach for vision-based lane-markings tracking. A robust and fast detection method is used to initialize the location of the lane-markings in the first frame of consecutive video sequences, and then process the lane-markings tracking phase. Due to the little change of the lane-markings location between consecutive frames in high sampling rate, Kalman filter is utilized to predict the lane-markings location of the next frame, and update the predicted lane-markings location via simple lane boundary detection. Therefore, the trend of lane-markings should be tracked robustness in the same frame and be applied to real-time intelligent vehicles. The experiments show that the proposed method is satisfactory to shadowy, non-continuous and continuous with curvature lane-markings.
Robust lane-markings tracking algorithm based on boundary detection and Kalman filter
2008
4 Seiten, 12 Bilder, 7 Quellen
Conference paper
English
Real-Time Robust Lane Marking Detection and Tracking for Degraded Lane Markings
British Library Conference Proceedings | 2017
|Real-Time Robust Lane Marking Detection and Tracking for Degraded Lane Markings
SAE Technical Papers | 2017
|The Detection of Lane Boundary Markings Using the Modified Spoke Filter
British Library Conference Proceedings | 1995
|Robust lane markings detection and road geometry computation
Springer Verlag | 2010
|Robust lane markings detection and road geometry computation
Online Contents | 2010
|