Autonomous Aerial Image Georeferencing for an UAV-Based Data Collection Platform Using Integrated Navigation System

Abstract

This research discusses the development of an automatic and no ground control points needed aerial image georeferencing method for the unmanned aerial vehicle (UAV) image data acquisition platform using a low cost integrated navigation system. An Extended Kalman Filter (EKF) based UAV navigation system was designed and implemented to fuse measurements from the inertial sensors (accelerometer and gyro), magnetometer and GPS. The navigation system was capable of providing continuous estimates of UAV position and orientation at 50 Hz even in the case of GPS signal outage. Based on the navigation data, the image collected by an onboard multispectral camera was automatically georeferenced and the ortho-rectified image was produced. Compared to eight presurveyed ground reference points, image automatic georeferencing results indicated that total position errors were less than 40 cm. This accuracy is sufficient for most of precision agricultural applications.