Research for Efficiency Improvement of Daily Inspection by Road Feature Extraction Using YOLO

Abstract

In the road maintenance, daily patrols are carried out to visually inspect the roads. However, it is difficult to detect a comprehensive change in the road surface by visual inspection. Therefore, we focus on the video data from the drive recorder installed in the patrol vehicle to support the daily inspection. In this paper, we propose a system that automatically recognizes road features from video data using YOLO.