Speed-Bump Detection for Autonomous Vehicles by Lidar and Camera

In this paper, we propose a speed-bump detection method for an autonomous vehicle by using a camera and light detection and ranging (lidar). A speed bump may have an impact with a vehicle if its speed does not decrease during driving. To prevent this, it is necessary to detect a speed bump and determine its position. In this study, we use a camera and lidar to detect and locate a speed bump. In addition, two detectors are used to extract and verify candidates for speed-bump. The detection method first extracts the regions of the speed-bump candidate using an image pattern. Then, using the image pattern and distance information, the speed bump is detected in the candidate area. The result includes the area of the speed bump, classification result, and speed-bump height information. The experimental results show that the proposed method improves the accuracy of detection and improves the classification accuracy of pedestrian crossings with similar patterns.