Self-positioning of a moving robotJazbar, Matic (Avtor) Vulić, Milivoj (Mentor) Pozicioniranjenotranji urezprojektivna geometrijarobothomogene koordinatefotogrametrijaCollinsova metodaExact knowledge of the position of a vehicle is a fundamental problem in mobile robot applications. A variety of systems and applications have been already been developed for mobile positioning on Earth, but not many of them are suitable for positioning on the Moon. Those methods are based on estimating the position of a robot by integrating an absolute and relative measurementsabsolute and relative measurements. Just like at indoor positioning also at With positioning on the Moon, just like with indoor positioning, there is no Global Navigation Satellite System (GNSS) available. In this case the idea of fusing several systems together becomes viableactive. On one side hand there are relative position measurement systemss such as odometry and inertial navigation and on the other handside there are absolute position measurements such as map matching, landmark based localization and active beacons. For integrating those sensors, a Kalman filter (EKF) is usually is used. The method presented here could be used for positioning a robot indoors or on the Moon. It integrates data gathered from a laser tracker and camera, and then calculates the position of the robot based on the Collins resection method. Data gathered from the experiment and possible locations of the robot are presented at the end. In mathematical calculation, the photogrammetry method and projective geometry were used, to avoid trigonometric and inverse trigonometric functions. In conclusion, all radial and tangential distortions were removed from the images and with the help of the iterative method the smallest deviations were calculated.20182018-09-12 07:46:34Magistrsko delo/naloga102935sl