Misalignment estimation of a lightweight laser scanning system with static calibration

Abstract

Abstract. Lightweight aerial laser scanning systems have become a practical alternative over the past few years for collecting 3D geospatial data, typically carried by drones. The positional accuracy expected for the collected data depends on the magnitude of errors generated by the sensors during data acquisition and data processing. These lightweight systems operate in kinematics mode, utilising less accurate attitude sensors compared to conventional aerial systems. Errors resulting from GNSS/IMU solutions (position and attitude) and boresight misalignment angles can significantly affect the accuracy of the point cloud. To reduce errors in platform positioning and other sources of error, a static calibration technique for lightweight systems is proposed in this paper. The technique is based on a static system assembly in a calibration field, using the displacement of a set of specific targets in the object space to simulate a flight path and significantly reduce errors from the positioning and attitude system. This is achieved by levelling the platform using the accelerometers by minimising accelerations in the X and Y axes. The misalignments of boresight angles are estimated based on observations at control points. A technique using the concept of Virtual Control Point (VCP) is also applied to reduce measurement errors. An experimental feasibility study was conducted using a system comprising an IbeoLux laser scanner and a NovAtel SPAN-IGM-S1 inertial measurement unit. The results showed that the technique is viable, but requires some improvements, mainly in using larger ranges. Improvements in boresight estimation were observed when using VCPs, compared to conventional calibration, especially in planimetry.