Automatic Detection of Damaged Roads and Lane Detection using Deep Learning

Abstract

This project introduces an automated system for detecting road surface damages and identifying lane markings using Deep Learning, YOLO (You Only Look Once), and Canny edge detection. The main goal is to improve road safety, assist autonomous navigation, and support efficient infrastructure maintenance. Road damages, such as potholes and cracks, are detected in real-time from images or videos captured by cameras mounted on vehicles or drones. The YOLO algorithm is used to classify and localize these damages with high speed and accuracy. At the same time, the Canny edge detection method identifies lane boundaries, ensuring precise lane detection even in challenging environments. Combining these techniques results in a reliable and scalable solution for smart transportation systems. The system reduces the need for manual road inspection and enables authorities to prioritize repairs based on real-time information. It also supports safer navigation for autonomous and assisted vehicles.