Abstract
Mostly, the Unmanned Air Vehicle (UAV) systems take the global position information from GNSS. In case of GNSS-denied environment, to perform autonomous flight mission and especially landing missions, it is necessary to know the position information of the air vehicle. In this study, a vision-based local positioning algorithm based on fiducial markers for Vertical Takeoff and Landing (VTOL) UAV systems is developed with a camera and markers. The corner points are obtained from the contour information of the markers in the images taken from the camera. Then, the position of the vehicle relative to the marker is calculated if the template images is matches with marker in the images from camera. After calculating the local position using vision-based algorithm, a one-dimensional Kalman filter is applied. Then, the local position is used to land autonomously. In addition, the local positions are calculated with different fiducial marker systems. The analysis for the application is performed in the simulation environment. Also, the experimental studies are conducted for accurate estimation and stable control of the UAV. Quadcopter physics is modeled mathematically in a simulation environment. The accuracy of the proposed vision-based position estimation algorithm is compared with the positions of the quadcopter in the simulation environment and the manual measurements in the experimental setup. The results demonstrated that the UAV position estimation is in an acceptable range.