Design of a Real-time Detection System for Potholes and Bumps Using Deep Learning

Potholes and bumps on asphalt roads pose daily hazards to vehicles, leading to accidents, crashes, and vehicle damage. Often, these damages occur when vehicles inadvertently speed over such irregularities. This study aims to address these issues by developing an object detection system that uses YOLOv8 to detect and classify potholes and bumps. A custom dataset was compiled by us for this purpose, then a model was trained on this dataset, and finally, real-time object detection was implemented. The system has been tested, and the results demonstrate its potential to be integrated into modern vehicles.