Developing an Image Processing Based Method for Rail Tracking with an Unmanned Aerial Vehicle in a Simulation Environment

Abstract

Rail transport is considered one of the safest means of transport in the entire world. With the development of high- speed trains, great investments have been made in our country in the field of railway transportation in recent years. With the increasing railway line, the demand for railway transportation is increasing day by day. Millions of people who are dependent on this means of transportation every day frequently use the railways. A preventive system is needed to keep up with this intensity, prevent unexpected breakdowns and support railway infrastructure systems. For the railway infrastructure to provide healthy service, maintenance services must be carried out regularly. Traditional railway maintenance services require intensive labor and time. This manual maintenance service is carried out with various rail maintenance tools. As an alternative to traditional methods, an unmanned aerial vehicle-based maintenance system that autonomously monitors and takes images of railway tracks is proposed in this study. The proposed method detects the vanishing point with the front camera of the drone. Then the PID controller is used to maintain the vanishing point tracking. The proposed method was developed in the Gazebo environment. The general purpose ofthe study is to record the rail images with the camera of the unmanned aerial vehicle that autonomously follows the vanishing point. In this way, real-life umnanned aerial vehicle experiments will be completed in a shorter time. © 2021 IEEE.