Assessment of an onboard EO sensor to enable detect-and-sense capability for UAVs operating in a cluttered environment

In an increasingly complex environment crowded with obstacles, particularly manned and unmanned traffic, technological advancements can autonomously provide alerts to the presence of incoming threats. In other words, advancements such as computer vision (CV) capability enhance overall situation awareness. This thesis explores the development and integration of CV capability onboard a functional unmanned aerial vehicle (UAV) to detect and track multiple proximate moving targets autonomously. A systems engineering approach is applied to define, analyze, and synthesize systematically a proposed system architecture for the real-time autonomous detection and tracking capability via visual sensors onboard the UAV. Both the hardware and software architecture design are discussed at length. Then, a series of tests that were conducted progressively to assess and evaluate the overall system architecture are described. Multiple UAVs and unmanned ground vehicles represented the contested operational environment. The developed CV algorithm proved successful at detecting and tracking multiple moving targets in real-time operation, thus laying the foundation for future research and implementation of the developed techniques in the automatic vision-based collision-avoidance guidance architecture.http://archive.org/details/assessmentofnonb1094556165Outstanding ThesisCaptain, Singapore ArmyApproved for public release; distribution is unlimited