基于HIS和小波变换的遥感图像云雾去除的研究  

Research on cloud removal methods of remote sensing image based on HSI and Wavelet Transformation

作　　者：张晖[1] 袁雪[1] 李海涛[2] 燕琴[2] 

机构地区：[1]北京交通大学电子信息工程学院,北京,100044 [2]中国测绘科学研究院 摄影测量与遥感研究所,北京,100830

出　　处：《新型工业化》2013年第5期-,共7页The Journal of New Industrialization

基　　金：教育部博士点基金

摘　　要：由于气象条件的不同，在实际的卫星遥感成像过程中，区域性的云雾是经常存在的。它们使得图像模糊不清，降低了图像的理解度和分辨率。而遥感图像中的云雾具有高亮度、低对比度、低频的特点。因此为了有效去除云雾噪声，本文将原始彩色遥感图像转化到HSI颜色空间，通过考察图像频率分布信息，结合小波变换的特点，本文仅对代表亮度的I分量进行小波变换。然后用新I分量与原始的H分量和S分量进行HSI反变换得到去云后的遥感图像。结果表明基于 HSI 和小波变换的云雾去除方法的去云影像，细节信息丰富，去云效果较好。Because of the different weather conditions, in the actual process of satellite remote sensing imaging, regional clouds often exit. They make the image blurred and reduce the degree of understanding and resolution of the image. The clouds on the remote sensing images have the characteristics of high brightness, low contrast, low frequency.so in order to effectively remove the cloud noise, the original color remote sensing image was transformed into the HSI color space. By examining the distribution information of the image frequency and combining the features of wavelet transformation, this paper only make wavelet transformation to I component of HSI color space. Then conduct HSI anti-transformation by combining the new V component with the original H component and S component of HSI color space to get processed results. The results indicate that the cloud removal method based on HSI color space and wavelet transformation could maintain the details of the image and achieve better results.

关 键 词：遥感图像 HSI颜色空间 小波变换 云雾去除 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术] P40[天文地球—大气科学及气象学]