基于混合Contourlet变换和主成分变换的高光谱图像融合

针对高光谱图像波段多、存在噪声干扰等特点,提出了一种混合Contourlet和主成分分析(principal com-ponent analysis,简称PCA)变换的高光谱图像融合方法.首先将多个波段的高光谱图像进行Contourlet变换,得到系列多尺度、方向各异的子带系数,然后利用PCA变换对各子带系数分别进行自适应融合处理.实验结果表明该算法可以有效地进行高光谱图像融合,消除噪声干扰,获得比直接应用Contourlet变换和PCA变换更好的融合效果.

Image Fusion of Hyperspectral Images Based on Hybdd Contourlet-PCA Transform

According to the characteristics of hyperspectral images such as muhi-band and corrupted by various noises, a nov-el fusion method of hyperspectral images based on hybrid Contottrlet transform and principal component analysis(PCA) transform were proposed. Firstly, Contourlet transform was performed simultaneously on several hyperspectral images to retain the sub-band images at different frequency and directions. Then each sub-band image was adaptively fused by PCA transform. The experiment result shows that the proposed method could well fuse hyperspectral images with noises eliminated and it outperforms the Contourlet and PCA methods.