Unmanned aerial vehicle relative navigation in GPS denied environments

This paper considers the problem of target handoff between Unmanned Aerial Vehicles (UAVs) in a GPS denied environment, and focuses on the design and evaluation of an estimation strategy for determining the relative pose of the aircraft. The estimation approach presented in this paper has three distinct components that act in concert to achieve the overall objective. First, a novel cooperative control and estimation strategy is used to determine relative pose from IMU and peer-to-peer ranging radio data without any a priori knowledge of either aircraft's pose. Next, a relative pose measurement is calculated using extracted features from downward looking cameras on the two UAVs. The computer vision technique first uses an indexing scheme based on a hierarchical statistical model to determine which frames from the two cameras have overlapping coverage, aligns the overlapping frames, and then calculates the relative pose estimate. Finally, a nonlinear Kalman Filter, which has been initialized with the a priori solution from the initialization filter is used to estimate relative pose by predicting it through the integration of IMU data of both UAVs with measurement updates from peer-to-peer radio ranging radios, magnetometers and the computer vision estimates.