Skeyeball: real-time vision system for an autonomous model airplane

The paper deals about real-time vision system for an autonomous model airplane. Sk^ey^e ball is a radio-controlled aircraft developed to host research and educational projects in autonomous aviation. Sk^ey^e ball vision has the objective of developing a tracking system that flows a simple two dimensional scene, in live video. It uses a two-prolonged approach to object tracking, taking into account the motion of the scene and the graphic signature of the object. Vision system is tested in the laboratory by using a rotating table, and in the air. The tracking system has been deployed in three sets of flights of the sk^ey^e ball aircraft, in the first two sets the pilot controlled the vehicle with verbal instructions from a ground operator monitoring the video feed. In the third, most recent set of flights is tested with a glimbal with increased range of motion, which was under manual control of the ground operator. The paper also presents the hardware architecture of the vision system and the airworthy implementation of the vision system both in the lab and the air.