Pothole detection system using 2D LiDAR and camera

Abstract

Automatic Pothole detection is important task for determining proper strategies of asphalt-surfaced pavement maintenance. In this paper, we develop a pothole detection system and method using 2D LiDAR and Camera. To improve the pothole detection accuracy, the combination of heterogeneous sensor system is used. By using 2D LiDAR, the distance and angle information of road are obtained. The pothole detection algorithm includes noise reduction pre-processing, clustering, line segment extraction, and gradient of pothole data function. Next, image-based pothole detection method is used to improve the accuracy of pothole detection and to obtain pothole shape. Image-based algorithm includes noise filtering, brightness control, binarization, addictive noise filtering, edge extraction, and object extraction and pothole detection. To show the pothole detection performance, experiments of pothole detection system using 2D LiDAR and camera are performed.