SafeDrive: Intelligent Pothole Detection and Mapping System

Potholes represent a substantial challenge impacting road safety and infrastructure durability, leading to increased vehicle wear, escalated maintenance expenses, and heightened safety risks. The SafeDrive: Intelligent Pothole Detection and Mapping System introduces an innovative, AI-enhanced methodology aimed at automating pothole detection and reporting. Leveraging YOLOv7, a cutting-edge object detection model trained using 11,068 images from the Roboflow dataset, including 1,808 validation images and 891 test images, the system processes video feeds to accurately identify potholes and geotag them for further analysis, achieving a precision of 87.72%, a recall of 77.03%, and an F1-score of 82.03%, ensuring effectiveness in real-world scenarios. Furthermore, the system categorizes detected potholes based on their occurrence frequency within video frames, enabling authorities to allocate resources efficiently by prioritizing critical areas. Additionally, it includes a user-friendly online application that allows users to report potholes with geotagged images and track repair progress, fostering community participation and enhancing transparency. Designed with scalability in mind, SafeDrive integrates cloud-based computing and sophisticated analytical tools to provide an efficient framework for road maintenance management, merging AI automation with active public involvement to significantly enhance road safety and set a new standard for monitoring urban infrastructure.