Pothole and Road Hump Detection Using YOLO Model

Every single one of the cars on the road, whether they're fully autonomous or not, relies on the stateof the roads to return safely. Tire wear and deadly traffic accidents are both caused by surfaceabnormalities like speed bumps and potholes. Thus, it is helpful to detect and characterize theseabnormalities in order to lessen the likelihood of accidents and vehicle damage. It is more difficult todiscover street anomalies in street photographs because of their inherent multivariate nature, whichincludes redundant and extensive information in addition to much tainted measurement noise. In thisstudy, we offer a YOLO-based automatic color picture analysis system for road pothole detection usingmobile camera snaps. A lightweight architecture was created to expedite both the training and usageprocesses. Each of the seven layers is appropriately aligned and responsive to one another. A pixel-forpixel reproduction of the original image is achieved. In an effort to gather as much data as possible, theconventional stride and pooling operations were administered. That makes the model we made better atspotting potholes and warning drivers to be extra careful.