a′Trous小波变换与PCA变换相结合的遥感影像融合分析  

Remote Sensing Images Fusion Based on a′ Trous Wavelet and PCA Transformation

作　　者：林卉[1] 景海涛[1] 张连蓬[1] 

机构地区：[1]徐州师范大学测绘学院,徐州221116

出　　处：《地球信息科学》2008年第2期269-272,共4页Geo-information Science

基　　金：江苏省高校自然基金项目(07KJD420215)

摘　　要：随着高分辨率遥感卫星的产生,传统的融合技术难以达到较好的融合效果,如主成分分析(PrincipalComponents Analysis)变换融合受到融合区域的限制,而传统的小波融合(Wavelet Transformation)算法由于高频直接替换,导致了一定程度的光谱失真,由此本文在分析主成分分析变换和a′Trous小波变换(WT)的基础上,以QuickBird全色和多光谱数据为实验数据,提出了一种将两者相结合的遥感影像融合方法,通过与其他融合方法的定量和视觉比较,结果表明该方法能得到更好的融合效果。As high-resolution remote sensing satellites come into being, it is hard to get good fusion results using original image fusion techniques. For example, the Principal Component Analysis （PCA） image fusion method is sensitive to the noise and limited to the fusion region. The traditional wavelet transformation fusion also causes spectrum information distortion to some extent, considering the effects of noise to the fusion image. Therefore, a＇ Trous wavelet and PCA transform integrated remote sensing fusion method is studied where Quickbird multispectral image and panchromatic image are used in this article. The result demonstrated that it is better than other fusion methods by visual and quantitative analysis.

关 键 词：主成分分析变换 a’Trous算法 小波变换 影像融合 

分 类 号：P237[天文地球—摄影测量与遥感]