• Version
• Download 21
• File Size 424.24 KB
• File Count 1
• Create Date 24 July 2025
• Last Updated 24 July 2025

Real-Time Lane Detection Using YOLO

ERUSU KATA RAJU REDDY, VEPADA VINOD KUMAR

Assistant professor, MCA Final Semester, Master of Computer Applications

 Sanketika Vidya Parishad Engineering College,

Vishakhapatnam, Andhra Pradesh, India.

ABSTRACT:

Lane detection is a challenging and long-standing problem in the field of computer vision. It involves complex visual cues and dynamic environments, making it a multi-feature detection problem that demands robust and efficient solutions. Traditional machine learning techniques have primarily focused on classification tasks rather than comprehensive feature extraction. While modern deep learning methods have demonstrated improved capabilities in feature detection, their application to efficient and accurate lane detection remains limited. In this paper, we propose a novel approach to lane detection by integrating advanced pre-processing techniques with a YOLO-based deep learning model. We first apply HSV color space transformation to enhance white and yellow lane markings and incorporate edge detection to strengthen structural features. Subsequently, a refined Region of Interest (ROI) selection is performed based on these features. The processed frames are then fed into a modified YOLO architecture tailored for lane detection, leveraging its real-time object detection strengths to identify lane boundaries effectively. Evaluations conducted on the KITTI road dataset demonstrate that our method outperforms existing pre-processing and ROI techniques, achieving superior accuracy and efficiency in lane detection tasks.

KEYWORDS: Lane Detection, Computer Vision, YOLO, Deep Learning, Edge Detection, Real-Time Detection,Object Detection,Pre-processing Techniques