为较好地保留多光谱、高分辨率遥感影像融合时的光谱信息和高分辨率信息,在小波变换基础上提出了一种基于模糊积分的融合算法,其基本思想是在IHS空间,对强度分量Ⅰ及高分辨率图像进行小波分解后,保留Ⅰ分量的低频系数,对于高频细节特征,将局部窗口内的方差、平均梯度和能量当作单因素指标,应用Choquet模糊积分综合各单因素指标得到一个综合指标,并据此来选取小波系数.实验结果表明,该算法在光谱质量改善方面明显优于IHS以及一些经典的小波变换遥感图像融合算法,是一种有效的遥感图像融合算法.

In order to decrease the spectral distortion while improving the spatial resolution during the fusion of remote sensing images, an image fusion method based on Choquet fuzzy integral is proposed in this paper. Firstly, an intensity-hue-saturation(IHS) transform is performed formultispectral images. Then the panchromatic image and the intensity component of multi spectral images are decomposed using discrete wavelet transform respectively. In thewaveletdomain, for the low-frequency componen,t the aveletcoefficients of the Icomponentare selected directly. For thehigh-frequency componen,t thewaveletcoefficients are selected according to an integrative index which is colligatedwith the values ofvariation, average gradientand energy in the local region by hoquet fuzzy integra.l Further, the fused intensity component is obtained by inverse discretewavelet transform. Finally, the fused images are obtained by inverse IHS transform. The experiment results demonstrate ourproposed method effectively.