基于TH-1卫星全色影像的多源遥感数据融合评价  

Image fusion quality assessment of multisource remote sensing for TH-1 PAN

作　　者：朱倩[1] 李霞[1] 刘玲玲[2] 

机构地区：[1]长安大学地球科学与资源学院,陕西西安710054 [2]陕西测绘地理信息局,陕西西安710054

出　　处：《中国科学技术大学学报》2013年第5期363-368,共6页JUSTC

基　　金：国家自然科学基金(41101445)资助

摘　　要：为寻求更加适合与天绘一号(TH-1)卫星全色影像融合的多光谱数据及融合方法,运用IHS,Wavelet,PCA,P+XS,Brovey 5种变换方法,分别将TH-1全色影像与TH-1,TM,ETM的多光谱影像进行融合,从视觉和定量两个方面对融合效果进行分析.结果表明:①3种融合组合方式中,TH-1PAN与TH-1MS的融合质量最高;②5种融合方法中,Wavelet变换光谱信息保持效果最佳,P+XS和Wavelet变换的空间信息保持效果较佳,Brovey和IHS融合的综合质量最佳.In order to find suitable multispectral images and fusion methods for the pan-sharpening of TH- 1 panchromatic images, IHS, Wavelet, PCA, P＋XS, Brovey five transformation methods were used to separately fuse TH-1 PAN with TH-1 MS, TM MS and ETM MS, and the fusion results were assessed from visual and quantitative aspects. Results show that TH-1 PAN and TH-1 MS fusion combination have the best quality; Wavelet transformation can keep the best spectral information, P＋XS and Wavelet transformation can keep the best spatial information, and Brovey and IHS can get the best overall quality.

关 键 词：TH-1 影像融合 融合质量评价 

分 类 号：TP75[自动化与计算机技术—检测技术与自动化装置]