Georeferencing of an Unmanned Aerial System by Means of an Iterated Extended Kalman Filter Using a 3D City Model

Abstract

Abstract In engineering geodesy, the technical progress leads to various kinds of multi-sensor systems (MSS) capturing the environment. Multi-sensor systems, especially those mounted on unmanned aerial vehicles, subsequently called unmanned aerial system (UAS), have emerged in the past decade. Georeferencing for MSS and UAS is an indispensable task to obtain further products of the data captured. Georeferencing comprises at least the determination of three translations and three rotations. The availability and accuracy of Global Navigation Satellite System (GNSS) receivers, inertial measurement units, or other sensors for georeferencing is not or not constantly given in urban scenarios. Therefore, we utilize UAS-based laser scanner measurements on building facades. The building latter are modeled as planes in a three-dimensional city model. We determine the trajectory of the UAS by combining the laser scanner measurements with the plane parameters. The resulting implicit measurement equations and nonlinear equality constraints are covered within an iterated extended Kalman filter (IEKF). We developed a software simulation for testing the IEKF using different scenarios to evaluate the functionality, performance, strengths, and remaining challenges of the IEKF implemented.