AI-Based Speed bumps, Potholes Detection and Alert System Using YOLOv8 Model.

Abstract: The rising incidence of accidents brought on by bad road conditions including potholes and unmarked speed bumps has raised serious concerns about road safety. These road irregularities frequently result in collisions and car damage, particularly when traveling at night or in bad weather. Road dangers cannot be tracked in real time using manual road inspection techniques, which are ineffective. In this research, a deep learning-based AI-based pothole and speed bump detection system is proposed. The system detects road dangers in real time using OpenCV image preprocessing and the YOLOv8 object detection model. The deep learning algorithm analyzes road photos taken by a car-mounted camera to find speed bumps and potholes. The technology informs the driver with both visual dashboard indicators and auditory cautions when it detects a hazard. Python, OpenCV, PyTorch, and Flask are used in the system's implementation. The suggested system can efficiently identify road risks and deliver real-time notifications, enhancing driver awareness and road safety, according to experimental results. Keywords: YOLOv8, Deep Learning, Computer Vision, Pothole Detection, Smart Transportation, Road Hazard detection