Robust mobile ground target localization using ground image features with UAV position compensation techniques

Recently, the task of localizing mobile ground targets using airborne images captured by unmanned aerial vehicles (UAVs) has received an increased interest of researchers in the UAV community. The task is even more difficult when position information of a UAV, estimated by Global Positioning System (GPS) signals, becomes unavailable. In this paper, we propose a new method to compute the UAV attitude and thus locate mobile ground targets using ground landmarks obtained from SIFT (Scale-Invariant Feature Transform) features. We show the validity of the proposed method using experimental flight data.