基于PCA 变换与小波变换的遥感图像融合方法

针对传统的PCA变换遥感图像融合技术会丢失部分多光谱遥感图像的光谱信息变量，从而造成光谱图像信息域的失真问题提出了基于PCA变换与小波变换的遥感图像融合方法。该方法首先提出多光谱遥感图像信息域的各波段相关矩阵的特征值变量和特征向量域，对多光谱图像进行主分量的变换，继而求得各主分量变量；然后将非灰度图像与多光谱图像信息域的首个主分量做直方图信息变量的匹配，利用小波变换融合方法来实现多光谱图像信息变量的首个主分量与非灰度图像的融合，其多光谱图像的首个主分量被融合结果来替代； 最后对多光谱图像信息变量的3个主分量变量作逆主分量变换得到所需的最终融合图像信息域。仿真实验表明，该方法使最终融合的图像在多光谱信息的保持与空间细节信息的增强两个方面的综合性能均得到提高。

Remote sensing image fusion based on PCA transform and wavelet transform

The traditional PCA image fusion can produce multi-spectral image information variable loss in remote image fusion. Aim to it, a new algorithm of remote sensing image fusion based on PCA and wavelet transform was proposed in this paper. Firstly, principal component transformation for multi -spectral image was performed by eigenvalues and eigenvectors in each wave band. Secondly, the first non principal component of non-gray image and multi spectral image were matched in histogram information. Finally, inverse PCA transform was carried out for three principal components to obtain the desired fusion image. Experimental results show the proposed algorithm does not only maintain multi spectral information but also enhanced the processed image details, and the processed image has better subjective visual effect and objective quantitative indicators.