Approaches to Autonomous Target Recognition

Unmanned aerial vehicles are becoming the defacto method of aerial surveillance for many applications. There is an ever increasing desire to fully automate these systems, from aerial navigation (autopilots) to sophisticated sensor systems that allow the vehicle to perform without a human operator. In this paper, we address the issue of target recognition using visual data to acquire a list of potential targets and then identify targets of interests and recognize their characteristics. When detailing recognition, we discuss two approaches and their implementations, as well as present a comparison of their results.