Real-time registration of airborne laser data with sub-decimeter accuracy

摘要

This paper presents a methodology for the precise registering of airborne laser data directly in flight with an accuracy that is sufficient for the majority of derived products, such as digital terrain models. We first present the strategy that integrates GPS/INS/LiDAR data for generating laser point clouds directly in flight and analyzes their accuracy. The latter requires the implementation of a functional covariance propagation on-line for all the system components (i.e. trajectory, laser, system calibration) to which influences of scanning geometry are added at the end of a flight line. The study of scanning geometry necessitates the classification of vegetation and coarse estimation of the terrain normal. This is achieved by a method that we formerly proposed for off-line quality analysis. The second part of the paper focuses on the positioning component. In high resolution scanning performed close to the terrain, the absolute accuracy of the resulting point cloud depends mainly on the quality of the trajectory which is related to the type of GPS solution (e.g. absolute positioning, DGPS, RTK). To reach sub-decimeter accuracy for the point cloud in the real-time, an RTK-GPS solution is needed. This requires the establishment of a communication link for the transmission of GPS corrections (or measurements). We analyze the usability of RTK-GPS/ALS acquired during several flights using different communication methods in the particular context of helicopter based missions. We focus mainly on the exploitation of nation-wide reference GNSS networks and confirm experimentally that the real-time registration of airborne laser data is feasible with sub-decimeter accuracy. Such quality is sufficient not only for a wide range of applications, but it also opens new opportunities for monitoring missions that require a short reaction time. Finally, we concentrate on situations when the phase and code corrections cannot be transmitted, and the quality of the differential carrier-phase positioning needs to be predicted. We validate the previously introduced indicators of positioning quality by simulated degradation of the input data.