• computer vision
• Image based visual servoing
• mobile robot
• opencv
• ros
• video stabilization
• video tracking
• visual path following

Mobile robotics is a research topic in which one or more robots can allow remote operations to a remote operator or move independently in an environment. One of the problems noted in teleoperation precisely concerns the presence of humans in the camera-robot system. It is imperative to optimize various criteria to enhance the efficiency of robot movements. However, it is crucial to recognize that these optimized paths may not always provide a visually comfortable experience for the human operator. The objective is to control a robot by allowing omnidirectional movements, in addition to the classic differential, following particular optimal trajectories, but making the experience to a teleoperator free of sensations such as nausea and discomfort during the autonomous movement of the robot. In this thesis the computation of an optimal path, in an obstacle-free environment, has been implemented, exploiting only the information coming from the camera, thus allowing the autonomous movement of the robot. Once the optimal path was calculated, an omnidirectional version was subsequently designed and implemented to follow the same visual trajectory as the differential mode. This design allowed for a comparison of different modes of movement, all based on the same visual path. Filtering of camera movements was finally implemented to make it smooth for a teleoperator to use it. The thesis ends with a critical examination of the achieved results. The practical tests carried out in the laboratory were like the simulations, however, with a large amount of oscillation produced. Also, the camera on the robot showed high latency and low framerate, Nevertheless, it was feasible to establish a direct comparison between the simulated and real-world scenarios.