卫星平台振动和反射镜震颤会引起遥感图像中的振荡畸变。这类畸变难以通过常用的几何校正方法消除。对此,提出了一种使用压缩感知的几何校正方法。该方法基于有理函数模型(RFM)进行几何校正。在校正过程中,利用初始的RFM计算出地面控制点(GCPs)在图像中的投影坐标与实际成像坐标之间的偏差(称为投影偏差),以地面控制点处的投影偏差作为采样值,使用压缩感知技术重构出所有像元处的投影偏差,并据此对RFM进行像方补偿;利用经过补偿的RFM进行遥感图像纠正。通过补偿,消除了振荡畸变引起的RFM模型误差,进而提高校正性能。利用实测数据验证了该方法的有效性,并通过仿真数据分析了地标点的数量与分布对该几何校正方法性能的影响。 Satellite platform vibration and mirror chatter can cause oscillations in remote sensing images. This type of distortion is difficult to eliminate by common geometric correction methods. In this regard, a geometric correction method using compressed sensing is proposed. The method performs geometric correction based on a rational function model (RFM). During the calibration process, the initial RFM is used to calculate the deviation (referred to as the projection deviation) between the projection coordinates of the ground control points (GCPs) in the image and the actual imaging coordinates. Taking the projection deviation at the ground control point as the sampling value, Reconstruction of projection deviations at all pixels using compressive sensing techniques, and based on which RFM is compensated for imagery; Remote sensing image correction using compensated RFM. By compensation, the RFM model error caused by oscillation distortion is eliminated, thereby improving the correction performance. The validity of this method is verified by the measured data, and the influence of the number and distribution of landmark points on the performance of the geometric correction method is analyzed by simulation data.