AI BASED ROAD CONDITION MONITORING POTHOLE DETECTION USING YOLOV11 AND DEEP LEARNING WITH PYTHON

Abstract

The Smart Road Monitoring System for AI-based pothole detection was developed to provide an automated and efficient solution for identifying road surface damages and assisting road maintenance authorities. Traditional road inspection methods require manual observation, which is time-consuming, labor-intensive, and often inaccurate. The maintenance of road infrastructure is a critical aspect of urban and highway management, directly impacting safety, vehicle longevity, and overall transportation efficiency. Traditional methods of road inspection, which rely heavily on manual surveys, are time-consuming, labor-intensive, and prone to human error. The proposed system addresses these limitations by using modern computer vision and deep learning techniques to automatically detect potholes from road inspection videos. To address these limitations, this project proposes an AI-based automated pothole detection and reporting system that leverages deep learning, computer vision, and automated reporting techniques to modernize road condition monitoring.The system utilizes a YOLOV11-based object detection model to analyze video footage captured from dashcams or road inspection cameras. Each frame is processed to identify potholes in real-time, with bounding boxes drawn around detected potholes and confidence scores calculated. A de-duplication logic ensures that each pothole is counted only once across video frames, enabling accurate identification of unique potholes. For every detected pothole, the system estimates its length, width, and depth, calculates volume, and determines material requirements including sand, cement, aggregates, and coal tar, as well as labor costs.The project also incorporates automated generation of professional PDF reports using the Report Lab library.