资源三号卫星影像融合算法对比分析  

Contrast and analysis of different fusion algorithms for ZY-3 satellite images

作　　者：王华斌[1] 李国元[1] 张本奎[2] 窦显辉 

机构地区：[1]国家测绘地理信息局卫星测绘应用中心,北京100830 [2]中国科学院研究生院,北京100190

出　　处：《测绘科学》2015年第1期47-51,共5页Science of Surveying and Mapping

基　　金：科技部科技支撑项目(2011BAB01B05)

摘　　要：我国第一颗高精度民用立体测绘卫星——资源三号卫星搭载了2.1m分辨率的全色相机和5.8m分辨率的多光谱相机,针对资源三号影像研究有效的影像融合算法对于提高资源三号的影像质量和应用范围具有很大意义。文章分析了Pansharp、Gram-Schmidt、Modified IHS、改进的SFIM和Brovey 5种融合算法,梳理了影像融合的质量评价指标体系,采用资源三号卫星影像覆盖的农田、城区、林地、裸地等不同类型的区域开展几种融合算法试验,并对试验结果进行了分析评价。实验结果表明,5种融合算法中,Pansharp融合算法对资源三号影像数据融合效果最优,Gram-Schmidt融合算法次之,其次为Modified IHS影像融合算法,改进的SFIM融合算法和Brovey融合算法融合效果较差。ZY-3satellite,the first civilian stereo mapping satellite with high precision in China,carries panchromatic camera with a spatial resolution of 2.1 meter and multi-spectral camera with a spatial resolution of 5.8meter.Searching for an effective image fusion algorithm for ZY-3satellite images is significant on the improvement of its image quality and application range.In this paper,five image fusion algorithms（Pansharp,Gram-Schmidt,Modified IHS,modified SFIM,Brovey）and different quality evaluation items were introduced.Finally,image fusion experiments on different areas with farm,city proper,forest land and bare land were implemented,and the result of image fusion was analyzed and discussed.It was showed that the five algorithms were listed in descending order by the fusion effects of ZY-3satellite images：Pansharp,Gram-Schmidt,Modified IHS,Modified SFIM,Brovey.

关 键 词：资源三号 融合算法 评价分析 

分 类 号：P228[天文地球—大地测量学与测量工程] TP75[天文地球—测绘科学与技术]