Lane Departure Warning System for Advanced Drivers Assistance

Lane detection is the important aspect of for the development of lane departure system. The lane detection is a very difficult task because it is affected by the different environmental condition. In this paper, effective computer vision algorithms is used to present the lane detection and lane departure system to avoid the road accidents. This method detects the edges of road and divider using canny edge detection algorithm while the edges are refined using Hough transform technique. The distance of vehicle from the middle divider is calculated by Euclidean distance. The Euclidean distance is combined with the PLSF algorithm gives the higher accuracy under different lighting conditions. This system is implemented using python language and OpenCV image processing libraries. Also, it is tested in real time environment using Raspberry Pi3 hardware module. The proposed system is efficient for straight and curve road as well as different lighting conditions.