遥感成像耦合典型误差源反演方法研究

根据遥感成像耦合典型误差源反演方法得到的误差源及其误差量,能够提高光学系统成像质量,使得高质量的遥感图像在军事和民用领域发挥出更大的作用。基于失真遥感图像,本文提出了一种遥感成像耦合典型误差源反演方法。首先根据典型误差源的调制传递函数模型和耦合误差源的解耦原理,构建耦合典型误差解耦方程组。然后利用刃边法提取遥感图像的调制传递函数,基于长短期记忆网络确定耦合典型误差源的初值。最后利用LM优化算法对耦合典型误差解耦方程组求解计算,反演出典型误差源及其误差量。实验结果表明,本文方法对失真遥感图像的耦合典型误差源的反演值与真实值的相对误差最大不超过8%,大部分都在5%以下,具有优异的反演性能。

Research on the inversion method of coupled typical error source in remote sensing imaging

Aceording to the eror source and the error amount obtained by the remote sensing imaging coupled typical error sourceinversion method, the imaging quality of the optieal system could be improved. The high-quality remote sensing image plays a great rolein the military and civilian fields, Based on the distorted remote sensing image, this study proposes a method for the coupled typical erorsouree inversion of remote sensing imaging. Firstly, according to the modulation transfer funetion model of the typical error source andthe decoupling principle of the coupled error source, the coupled typieal error decoupling equations are formulated. Then, the modulationtransfer function of remote sensing image is extracted by the edge method. The initial value of the coupled typical error source isdetermined based on the long-short term memory network. Finally, the Levenberg-Marquardt optimization algorithm is used to solve andcaleulate the coupled typical error decoupling equations. The typical error source and error amount are reversed. Experimental resultsshow that the relative eror between the inversion value and the true value of the coupled typical error of the distorted remote sensingimage is no more than 8%,most of which are below 5%. The method has excellent inversion performance.