RDD: ROAD DAMAGE DETECTION AND CLASSIFICATION USING YOLOV8

Road damage, such as cracks and potholes, poses serious transportation safety risks and exponentially increases maintenance costs if not detected promptly.Current road inspection methods-manual surveys-are expensive, time-consuming, and subjective.This paper introduces RDD: Real-Time Road Damage Detection System, an automated solution leveraging YOLOv8, a modern, high-performance deep learning object detection framework.The proposed system surpasses traditional methods by performing direct, single-stage object detection to concurrently identify the type and precise location of damage (including multiple crack types and potholes) from both captured images and real-time video streams.The system is engineered for deployment using vehicle-mounted cameras.Experimental results demonstrate high accuracy, reliable real-time performance, and robustness across varying environmental conditions.This makes RDD an ideal and practical solution for deployment in intelligent transportation systems and smart city infrastructure monitoring.