Abstract: This paper discusses the PCA（Principal Component Analysis） and a few remote sensing image fusion methods that combine the PCA and the WT（Wavelet Transform）.Traditional PCA algorithms loss some useful information about spectral characteristics of the first principal component of multi-spectral images，and influence largely the spectral resolution of fusion images.A remote sensing image fusion algorithm is implemented based on the feature product of the PCA and wavelet，which makes use of the time-frequency analysis features of wavelet transform and feature product that fuses the first principal component of multi-spectral images.The improved algorithm has been tested on various multi-spectral and panchromatic satellite remote sensing images for different areas.Experiment results show that the algorithm has a better fusion result from the perspective of either subjective visual effect or objective statistical data.

Key words: feature product, Principal Component Analysis（PCA） transform, remote sensing image fusion, wavelet transform

摘要： 在遥感图像融合中，传统PCA算法会损失部分有用信息，从而使得融合结果的光谱分辨率受到较大影响，针对这种情况，借助小波变换优良的时频分析特性，利用特征量积来融合多光谱图像的第一主成分，实现了一种基于特征量积与PCA的小波遥感图像融合算法。通过对来自不同场景不同卫星的多光谱和全色图像进行融合实验，结果表明，该算法无论在主观视觉还是在客观统计数据上，均具有比其他方法较佳的融合效果。

关键词: 特征量积, 主成分分析（PCA）变换, 遥感图像融合, 小波变换