遥感图像薄云去除方法及IDL的实现研究  

Research on the haze removal method of remote sensing images and the implementation of IDL

作　　者：刘建[1,2] 何政伟[1,3] 杨斌[4] 李何超[2] 张东辉[1] 刘敏红[1] 

机构地区：[1]成都理工大学地质灾害防治与环境保护国家重点实验室,成都610059 [2]四川省国土勘测规划研究院,成都610031 [3]首都师范大学资源环境与地理信息系统北京市重点实验室,北京100037 [4]西南科技大学环境与资源学院,四川绵阳621010

出　　处：《测绘科学》2010年第5期77-79,共3页Science of Surveying and Mapping

基　　金：国家十一五科教支撑计划(2008BAK49B02);国家863重点项目(2007AA120306)

摘　　要：云噪声的干扰,严重影响了遥感图像的解译与信息提取。本文从遥感图像的成像机理出发,使用同态滤波算法,采用IDL二次开发语言并集成到ENVI平台上,实现了遥感图像薄云的自动化检测与去除。实验证明,该过程能够有效地去除薄云,提高在传统方法中图像预处理的效率,在一定程度上提高了图像的质量与精度。The interference of the cloud noise affects seriously the interpretation and information extraction of remote sensing images.Based on the imaging mechanism,according to homomorphic filtering algorithm,by using IDL redeveloped-language and integrating it to the ENVI platform,this paper implemented automatic detection and removal of haze in remote sensing images.Experimental results showed that this process could effectively remove the haze and efficiently improve the traditional methods of images＇ pre-processing.In addition,it is improved the quality and precision of the images greatly.

关 键 词：薄云去除 IDL 同态滤波 遥感图像 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]