基于区域特征加权的IHS图像融合方法

An IHS Image Fusion Method Based on Weighted Regional Features

针对传统IHS变换图像融合方法在高分辨率和多光谱图像融合中存在的光谱畸变现象,研究了现有的一些改进IHS变换图像融合方法,提出了一种基于局部区域多元特征动态加权的IHS图像融合方法,根据主观目视判决和客观评价指标对融合结果进行了比较和分析。仿真结果表明,该文提出的方法能使融合图像具有较好的空间分辨率,同时又能很好地保持原有多光谱图像的光谱特性。

IHS (Intensity-Hue-Saturation)transformation is one of the most commonly used method while fusing high res-olution image and multispectral image,but traditional IHS transformation method faces a severe problem namely color distortion.This paper reviews several improved IHS transformation image fusion algorithm,and proposes a new fusion method,which fuses the high resolution gray image and the Intensity component of the multispectral image throuth adaptive weights based on window region features such as entropy and average gradient.Experimental results prove the effectiveness of the proposed fusion method.