高分影像辅助下的航空高光谱严密几何检校方法

 惯性测量单元(IMU)与传感器视准轴的偏心角和偏心矢量是造成航空线阵列高光谱数据几何校正误差的主要原因之一。在分析偏心角与偏心矢量误差来源之后提出该误差由IMU主轴与传感器主轴的角度偏差、测区固定偏差、GPS中心与传感器投影中心相对偏差组成,在此基础上建立了较为严密的检校模型。针对模型解算时需要大量高精度控制点的问题,提出了一种高分影像辅助下的亚像元精度控制点自动提取方法。通过多地区、多传感器高光谱航测实验表明,亚像元精度控制点能有效提高模型解算精度。新检校模型可获得亚像元校正精度,推扫式传感器——应用型机载成像光谱仪(AISA)建模中误差约为0.39个像元,摆扫式传感器——实用型模块化成像光谱仪(OMIS)建模中误差约为0.23个像元,校正后的影像可直接进行拼接。

A strict geometric calibration method for airborne hyperspectral sensors aided by high resolution images

The eccentric angle and eccentricity vector between inertial measurement unit (IMU) and the collimation axe of sensors are the main causes for geometric correction errors of airborne line array hyperspectral Images. In this paper, the errors caused by eccentric angle and eccentricity vector are supposed to consist of the angular deviation between spindles of IMU sensor and hyperspectral camera, fixed position offset, relative position offset between GPS center and sensor projection center firstly, and then a strict geometric calibration model is given. As this new model needs a large number of high-precision control points for calculation, an automatic control point selection method with sub-pixel precision is proposed. Experiments of multi-region and multi-sensor show that sub-pixel control points selection method can improve the accuracy of the calibration model effectively. With this new calibration model, the push-broom sensor (airborne imaging spectroradiometer for applications (AISA)) modeling error is about 0.39 pixels, and the whisk-broom sensor operational modular imaging spectrometer (OMIS) modeling error is about 0.23 pixels. Both of the geometric correction precisions are at sub-pixel level. The images after geometric correction can directly mosaic together.