Enhanced Lane Detection Techniques on Structured Roads using OpenCV

Lane detection remains a significant challenge in Advanced Driver Assistance Systems (ADAS) due to its susceptibility to various factors such as complex road patterns, adverse lighting and weather conditions, and the presence of obstacles like shadows and other vehicles. To address these vulnerabilities, this paper proposes a real-time lane detection approach utilizing the Sliding Windows method and the Canny edge detection technique implemented with the OpenCV Python package. The proposed method includes distortion correction, preprocessing, lane detection on perspective-transformed frames, and lane visualization as output. Additionally, the method calculates the lane’s radius of curvature and the car’s offset from the lane center. In the preprocessing stage, several edge detection methods, including Laplacian, Sobel, and Canny, are evaluated. Through rigorous comparison and analysis, the superiority of the Canny edge detection method is demonstrated. The performance of the proposed approach is evaluated on test videos obtained from the Udacity GitHub repository, showcasing its effectiveness in challenging scenarios.