Master’s Thesis for the graduation in Artificial Intelligence

ii The small size of micro aerial vehicles (MAVs) allows a wide range of robotic applications, such as surveil-lance, inspection and search & rescue. In order to operate autonomously, the robot requires the ability to known its position and movement in the environment. Since no assumptions can be made about the en-vironment, the robot has to learn from its environment. Simultaneous Localization and Mapping (SLAM) using aerial vehicles is an active research area in robotics. However, current approaches use algorithms that are computationally expensive and cannot be applied for real-time navigation problems. Furthermore, most researchers rely on expensive aerial vehicles with advanced sensors. This thesis presents a real-time SLAM approach for affordable MAVs with a down-looking camera. Focusing on real-time methods and affordable MAVs increases the employability of aerial vehicles in real world situations. The approach has been validated with the AR.Drone quadrotor helicopter, which was the standard platform for the International Micro Air Vehicle competition. The development is partly based on simulation, which requires both a realistic sensor and motion model. The AR.Drone simulation model is described and validated