Route detection and autonomous route tracking of unmanned aerial vehicles using image processing

In this study, a software has been developed for route tracking with Unmanned Aerial Vehicle (UAV) by using image processing algorithms. The recognition of the predetermined line by using image processing algorithms and the process of route-following by using the method of Tangent Vector Field Guidance (TVFG) have been performed in indoor and outdoor experiments. Route-following algorithm is investigated under the effects of wind and light levels that deflect the flight of UAV. In the tests, the quadrotor has been controlled by its 9-channel controller and the program that is developed with C#. After the images have been transferred to computer wirelessly from the camera attached to UAV and image processing has been completed, TVFG approach has been used. In TVFG, the direction angles calculated by using the vector fields that will follow the line-shaped guide path are used. When the path to be followed has more than one direction instead of a single straight line, it is divided into sections which consist of straight lines, and by prioritizing these lines, the most dominant line is followed. In this study, it has been elaborated to provide a dynamic model by considering the tracking errors. As a result of the process adopted, UAV's autonomous flight is achieved by using the visual inputs and TVFG method, and the external deflecting factors are investigated.