基于改进矩匹配算法的珠海一号高光谱卫星影像相对辐射校正

Relative Radiometric Correction of OHS Remote Sensing Images Based on Improved Moment Matching

珠海一号高光谱卫星影像部分波段存在着噪声干扰,严重影响了数据的质量与信息利用。本文分析了珠海一号高光谱卫星数据噪声的特点,在矩匹配算法的基础上,使用移动窗格对影像进行分段处理,并采用距离加权计算与去除极端值统计数据的方法提高参考值的准确度,实现了基于影像的相对辐射校正。实验结果表明,与传统矩匹配法和傅里叶滤波法相比,本文方法能有效去除遥感影像中存在的噪声,并能较好保留原影像的辐射特征。

Some bands of Orbita Hyper Spectral(OHS)images are corrupted by noises,which seriously affect the data quality and information utilization.This paper analyzes the characteristics of the OHS data noises.On the basis of moment matching algorithm,the image is processed with a moving-window method,using distance weighting calculation and removing extreme value statistics to improve the accuracy of reference value.It realizes the image-based relative radiation correction of the images.The experimental results show that this method can effectively remove the noises in the image,and the radiation characteristics of the original image is well retained,as compared with the traditional moment matching method and Fourier filtering method.