Precision Directly Georeferenced Unmanned Aerial Remote Sensing System: Performance Evaluation John Perry, Geomatics Program Ahmed Mohamed, Ph.D., Geomatics Program, School of Forest Resources and Conservation Amr Abd-Elrahman, Ph.D., Geomatics Program, School of Forest Resources and Conservation Scott Bowman, Department of Mechanical Engineering Youssef Kaddoura, Geomatics Program Adam Watts, Wildlife Biology University of Florida Gainesville, FL BIOGRAPHY John Perry is pursuing a Master Degree after graduating with a Bachelor Degree in Geomatics at the University of Florida. His background is in spatial measurement systems, with a research focus in Inertial Navigation Systems. Ahmed Mohamed is an Assistant Professor of Geomatics at the School of Forest Resources and Conservation at the University of Florida. He received his Ph.D. at the University of Calgary Geomatics Engineering Department. His research is INS/GPS integration as the basis for direct georeferencing on remote sensing systems. Amr Abd-Elrahman is an Assistant Professor of Geomatics at the School of Forest Resources and Conservation at the University of Florida. He received his Ph.D. from the University of Florida. His research work is in remote sensing and image processing. Scott Bowman is pursuing a Master Degree in Mechanical Engineering after receiving a Bachelor Degree in Electrical Engineering from the University of Florida. His research is in autonomous flight control systems. Youssef Kaddoura is a Ph.D. student at the University of Florida, and is currently working with synchronization mechanisms in remote sensing and georeferencing systems. Adam Watts is a Ph.D. student in Wildlife Biology at the University of Florida, and is investigating the Unmanned Aerial Remote Sensing System as a platform for biological and environmental surveys. ABSTRACT This paper presents a practical evaluation of a direct georeferencing implementation with a quantitative analysis of the error budget of the georeferencing solution with respect to an Unmanned Aerial Remote Sensing System (UARSS). Direct georeferencing of remote sensing data using on-board navigation sensors is a critical step toward making near real-time remote sensing possible. By eliminating the need for costly and time- consuming ground control points, direct georeferencing provides the ability to fly over areas that do not have to be surveyed beforehand. This can decrease the resources required to produce usable remote sensing data by orders of magnitude. Particularly suitable are natural resource, agricultural, and infrastructure monitoring applications, where rapid and repeatable acquisition of the data directly into a Geographic Information System would be advantageous. A serious limitation of these systems is providing sufficient accuracy using low-cost and low-weight navigation sensors. This paper discusses a program for assessing the accuracy of the navigational sensors, as well as the results of a real-world implementation. A benchmarking platform for comparing the UARSS sensors to high-accuracy sensors is used for a ground- based test. Misclosures evaluated for determining the overall accuracy of the direct georeferencing system. The components of the error budget are presented, with an analysis of those elements relevant to the scope of this paper. Methods are presented for evaluating these additional components. The design goal of one meter ground accuracy is not met, and a forward-looking plan is presented for improving the performance of the system. INTRODUCTION Unmanned Aerial System (UAS) technology has provided a platform that is highly mobile, relatively inexpensive, and safe to operate, opening up a huge potential for expanding the scope and availability of remote sensing data at biologically relevant scales [Pearlstine et. al. 2001]. Work in this area has resulted in the advent of the Unmanned Aerial Remote Sensing System (UARSS), a generic term for UAS systems with integrated remote