基于空间频率的图像融合方法

Image Fusion Based on Spatial Frequency

论文对多光谱图像和高分辨图像进行了融合。对多光谱图像进行IHS变换,利用变换后得到的强度分量和高分辨图像具有较强的相关性,在小波变换域进行图像融合。根据小波变换有三个方向的高频细节这一特点,提出了一种新的空间频率概念,基于这种空间频率进行图像融合,得到了同时具有较好的空间分辨率和光谱信息的融合图像。实验结果表明该方法得到的融合图像优于传统的IHS变换法和传统小波变换方法。

Multispectral and high resolution images fusion is presented.The high correlation between high resolution image and spectral intensity component by means of the IHS transform,makes it easy to perform images fusion in wavelet transform domain.A new spatial frequency is put forward by wavelet transform characteristic of having three orientations detail component.Experimental comparisons show that the method proposed is better than traditional IHS and wavelet transform fusion method in preserving spatial resolution and spectral component.