Self-calibration method for eccentric angle in hyperspectral and LiDAR integrated airborne system

Abstract

Eccentric angle between IMU and the collimation axis of airborne sensor is one of main reasons causing geometric correction error. Currently, the application of the hyperspectral and LIDAR integrated system are greatly affected by lack of universal calibration methods which can calibrate the hyperspectral data and LiDAR data simultaneously. An eccentric angle calibration model for hyperspectral and LiDAR integrated system is proposed, meanwhile a self-calibration method using “#” shaped flight strip is designed to validate the new calibration model. Firstly, the homogeneous points are searched from all geometrically corrected flight strips by automatic matching methods. Secondly, control points obtained by averaging the coordinates of the homogeneous points are applied to solve the calibration model to get the eccentric angle. Finally, the point clouds are corrected geometrically with the solved eccentric angle, then the above steps are repeated until the solved eccentric angles are stable. An experiment was carried to testify the new calibration model and resolving method, which shows that the proposed model is of high-precision, validity and the resolving method is of fast convergence. The airborne sensors can acquire a plane precision of 0.807 m (about 1.2 pixels) without any ground control points, and the LiDAR can acquire a plane precision of 0.437 m and elevation precision of 0.15m.