基于颜色空间变换的遥感图像去厚云方法  

Cloud Removal of Remote Sensing Imagery Based on Color Space Transformation

作　　者：陈旭[1] 林宏[1] 强振平[1] 

机构地区：[1]西南林学院计算机与信息科学系,昆明650224

出　　处：《遥感信息》2009年第3期38-43,共6页Remote Sensing Information

基　　金：西南林学院重点科研基金项目资助(项目编号:110712)

摘　　要：由于气候条件等因素,获取清晰无云的大面积遥感图像几乎不可能。在所获取的遥感图像中常含有大量的厚云完全遮盖了地表的实际地物情况。提出了一种基于遥感自动分类和颜色空间变换的多时相遥感图像厚云去除方法。实验结果表明,该方法不但能够去除厚云,而且能够很好地校正不同时相遥感图像间的颜色和亮度差异。Due to the weather conditions and other factors, the cloud-free access to clear large areas of remote sensing image is almost impossible. The acquired remote sensing images often contain large quantities of thick cloud covering all the surface fea- tures in the actual situation. This paper presents a thick cloud removal method of the multi-temporal remote sensing image based on automatic classification and color space conversion. The experimental results showed that this method can not only re- move the thick clouds, but also correct the color and brightness differences on different temporal remote sensing images.

关 键 词：遥感 色彩校正 去云 颜色空间变换 

分 类 号：TP751[自动化与计算机技术—检测技术与自动化装置]