Vision Based Tracking for Unmanned Aerial Vehicle

This project aims at a system for an Unmanned Aerial Vehicle (UAV) following moving targets on ground. The UAV has physical constraints on airspeed and maneuverability. The target however can move freely and in any general pattern. We assume minimum knowledge about the target while navigating the aircraft. The system includes visual tracking of the target with a camera mounted on the UAV. The camera is also controlled by the closed loop algorithm according to the position and orientation of the aircraft and the position of the target. Aircraft stabilization and interpretation is performed using an autopilot system "Piccolo" by Cloud Cap Technology. The navigation and visual processing is performed on computers at the ground, control commands from which are sent to the aircraft wirelessly. The figure shows our system architecture and pictures of the individual components. Secondly, the project aims at design and implementation of a real-time computer vision system for a rotorcraft unmanned aerial vehicle to land onto a known landing target. This vision system consists of customized software and off-the-shelf hardware which perform image processing, segmentation, feature point extraction, camera pan/tilt control, and motion estimation. We introduce the design of a landing target which significantly simplifies the computer vision tasks such as corner detection and correspondence matching. While UAVs have been widely used in the West for policing, oil pipeline inspection and detecting illegal immigrants, in India, they are mainly used by the armed forces (Eg. Lakashya, Nishant) for intelligence gathering and surveillance. But this is changing.