Lane and Object Detection using YOLO: Indian Roads Scenario

Abstract

This research explores the feasibility of You Only Look Once (YOLO), a deep learning object detection algorithm, for lane and object detection in challenging Indian road environments. Traditional methods struggle with faded lane markings, dense and diverse traffic, and unpredictable scenarios. YOLO's speed and accuracy make it suitable for real-time ADAS applications. The methodology leverages Roboflow, a platform for computer vision tasks, to explore data acquisition, model selection, training, and evaluation. This research aims to contribute to developing safer and more reliable ADAS systems for Indian roads.