Country Road Boundaries Detection for Autonomous Driving Based on LiDAR Point Cloud

Drivers still rely on their judgment and operation when driving on unstructured roads under poor lightening conditions due to the shortage of machine vision. Thus, there is a growing demand for improving the driverless technology on unstructured roads. In order to solve the conflict between cost of sensors and accuracy of detection results, this paper proposes a method that filtering non-road points and extracting road boundaries based on 16-beam LiDAR which is hard to be affected by the change of light. The proposed method was applied to three data sets collected from different driving circumstances in the countryside. The performance of the method was recorded respectively which showed the feasibility and efficiency.