一种基于ENVI二次开发的遥感薄云去除方法的改进

An Improved Approach to Remove Cloud in RS Image Based on the Second Exploiture of ENVI

近年来常用的同态滤波去云算法因其采用滤波器的局限性只能去除低频区域的云而无法去除高频区域的云。因此在分析传统的同态滤波去云算法的基础上,引入了空域滤波,采用中值滤波器对图像进行处理,旨在去除高频区域的云。而后在ENVI遥感图像处理软件平台中采用IDL语言实现了算法并对ENVI进行二次开发。经实验结果分析表明,该法是有效的,并且为遥感数据的后续应用提供了方便。

The traditional homomorphic filtering frequently used in cloud-removing can remove the cloud in low-fre- quency, but can＇t in high-frequency because of the limitation of the filter it uses. In order to improve it, median filter which is used to deal with the image in spatial is introduced at the basis of the study of the traditional algorithm. And then the algorithm is accomplished in IDL and ENVI is developed. At last analysis of the experiment in two methods is carried out, which shows the improved method both effective and convenient for the application of the mass RS data.