一种基于 Sobel 算子的小波与 IHS 遥感影像融合算法

摘要

针对IHS图像融合算法中颜色畸变比较明显的问题,提出一种新的基于小波与IHS相结合的遥感影像算法.经IHS变换的多光谱影像Mul的I分量与全色影像Pan由二维离散小波分解,对小波高频和低频分量采用不同融合规则:低频分量采用绝对值加权平均的方法,把两者的低频系数按其权值比例合成到新的分量I1中;高频系数采用基于区域分块的Sobel算子的绝对值取大. 实验结果与IHS法、传统小波与IHS结合法相比较,该算法能获取更多的光谱信息,人眼视觉效果也较好.%Aimed at IHS color distortion in image fusion algorithm is more noticeable problems ,put forward a kind of based on wavelet and IHS improved algorithm with the combination of remote sensing image .Multispectral images by IHS transform the Mul I component and panchromatic images of Pan by the two-dimensional discrete wavelet decomposition , the wavelet high frequency and low frequency components with different fusion rules:low frequency component of the ab-solute value of the weighted average method ,the low frequency coefficients according to the weight ratio of the two syn-thetic I1 to new components;High frequency coefficients based on region partition is the absolute value of Sobel operator take big.Experimental results with IHS method ,traditional wavelet combined with IHS method comparison ,the proposed algorithm can obtain more spectral information ,the human eye visual effect is better also .