GeoTrack: an autonomous closed-loop target tracking system for small UAV networks


This paper describes a closed loop tracking and control system for a collaborating team of Small Unmanned Air Vehicles (SUAVs). The Geotrack system includes automatic target detection and tracking with video, real-time video geo-registration, distributed sensor fusion, and bio-inspired collaborative UAV control. The nominal sensors employed by the system will be EO/IR video cameras on the SUAVs. Video tracking algorithms perform fully automated multipletarget detection and feature-aided tracking, geo-registration, and multi-sensor data fusion. Control algorithms are derived from swarming algorithms used by flocks of birds and schools of fish. These algorithms will enable SUAVs to move in formation and cooperatively track targets with minimal estimation errors, while relying solely on local communication between the agents. Implementation of the video tracking, compression, and geo-registration will be in a low-SWAP FPGA device that can run on-board a SUAV. This device and system is expected to improve the ISR capabilities and autonomy of SUAVs over the current state-of-the-art. This paper presents first-year preliminary progress and results.

Collins, G., Barooah, P., Hespanha, J.
Proceedings of the International Conference and Exhibition of Unmanned Aerial Vehicles
Date: June, 2010
ICB Affiliated Authors: João P Hespanha