Lane Detection Based on Road Module and Extended Kalman Filter
Lane detection is already a basic component in modern vehicle control systems, with satisfying accuracy on labeled roads. There are still occasional problems of low accuracy and robustness in cases of challenging lighting, shadows, or in cases that road marking is missing. The paper proposes a new algorithm combining a model of road structure with an extended Kalman filter. Lane borders are detected by an adaptive edge detection operator based on scan lines. A new parameter space is defined to adjust the algorithm to the current lane model. All candidate lanes are extracted by voting of edge points. Road boundaries are obtained by considering various constraints. A new driveway model is specified according to roadway geometry and vehicle dynamics. The estimation of parameters is expanded for also covering driveway information. Coordinates of lane border points are tracked and estimated using an extended Kalman filter. Special attention is paid to enhancing stability and robustness of the algorithm. Results indicate that the proposed algorithm is robust under various lighting conditions and road scenarios; it is also of low computational complexity.
This work is supported by National Natural Science Foundation of China (Grant No. 61471272), Natural Science Foundation of Hubei Province, China (Grant No. 2016CFB499).
- 3.Aly, M.: Real time detection of lane markers in urban streets. In: Proceedings of IEEE Intelligent Vehicles Symposium, pp. 7–12 (2008)Google Scholar
- 4.Xiao, J.-S., Cheng, X., Li, B.-J., et al.: Lane detection algorithm based on beamlet transformation and k-means clustering. J. Sichuan Univ. (Eng. Sci. Ed.) 47(4), 98–103 (2015). (in Chinese)Google Scholar
- 5.Meng, L.-X., Sun, F.-C., Shao, Y.: Survey on road image interpretation based on monocular vision. J. Comput. Appl. 30(6), 1552–1555 (2010). (in Chinese)Google Scholar
- 7.Yu, B., Jain, A.K.: Lane boundary detection using a multiresolution Hough transform. In: Proceedings of IEEE International Conference on Image Processing, pp. 748–751 (1997)Google Scholar
- 11.Zhou, S., Jiang, Y.: A novel lane detection based on geometrical model and Gabor filter. In: Proceedings of IEEE Intelligent Vehicles Symposium, pp. 59–64 (2010)Google Scholar
- 12.Mechat, N., Saadia, N., M’Sirdi, N.K., Djelal, N.: Lane detection and tracking by monocular vision system in road vehicle. In: Proceedings of IEEE International Conference on Image, Signal Processing, pp. 1276–1282 (2012)Google Scholar
- 13.Kortli, Y., Marzougui, M., Atri, M.: Efficient implementation of a real-time lane departure warning system. In: Proceedings of IEEE International Conference on Image Processing Applications Systems, pp. 1–6 (2016)Google Scholar
- 15.Lee, D., Shin, J., Jung, J., et al.: Real-time lane detection and tracking system using simple filter and Kalman filter. In: Proceedings of IEEE International Conference Ubiquitous Future Networks, pp. 275–277 (2017)Google Scholar
- 17.Hough, P.V.C.: Method and means for recognizing complex patterns. US Patent, pp. 77–79 (1962)Google Scholar
- 20.Watanabe, A., Naito, T., Ninomiya, Y.: Lane detection with roadside structure using on board monocular camera. In: Proceedings of IEEE Intelligent Vehicles Symposium, pp. 191–196 (2009)Google Scholar
- 21.Welch, G., Bishop, G.: An introduction to the Kalman filter. UNC, Chapel Hill, TR (2006)Google Scholar