摘要
通过遥感图像融合,可获得更丰富的信息,常用遥感图像融合方法如HIS彩色变换可将不同平台、不同光谱响应范围的高空间分辨率的遥感数据与多光谱遥感数据进行融合,但要求这两组数据的光谱响应范围一致,否则便会产生光谱扭曲的现象,从而影响了地物的识别。针对不同平台、不同光谱响应范围的遥感数据,该文在HSI变换的基础上,提出了一种改进的方法,即LFF融合法,首先对高几何分辨率的全色波段进行LoG滤波,而后将LoG滤波后的全色波段与多光谱经HSI正变换后的强度分量进行灰度直方图匹配,并替换之,经HSI逆变换便得到融合图像。论文从灰度变化指数和分类精度两方面分析了光谱保持性能和融合图像的分类精度,分析结果表明:LFF融合法的光谱保持性能优于HSI变换法,LFF融合后图像的分类精度高于HSI融合后的图像,LFF融合法是一种能较好地保持光谱特性的融合方法。
Fusion production of Remote sensing image may acquire detailed con te nt from multi-source RS data at less cost.These conventional fusion methods s uch as HSI transform can merger two optical image data of different resolutio ns,a high spatial resolution panchromatic image and a low spatial resolution bu t multi-spectral image.However,this fusion method requires same or analogou s spectral response range between the high spatial resolution panchromatic image and the low spatial resolution multi-spectral image.This paper brings forward an improved fusion method called LoG Filter Fusion(LFF)based on HSI transfor m,which can merge two different kinds of optical image data with different s pectral response range.LFF algorithm is put forward in this paper,which first ly filters the high spatial resolution panchromatic image with LoG filter,and t hen substitutes the intensity component through HSI transform,lastly produces the LFF image by HSI invert transform.The results show the gray variance inde x of LFF fusion method is higher than HSI transform,and the color of LFF image is analogous to the low spatial resolution multi-spectral image.So the LFF fu sion method is more excellent at the spectral preservation than HSI transform approach,and the classification accuracy of the LFF image is more5.69%high er than HSI image.
出处
《计算机工程与应用》
CSCD
北大核心
2004年第11期66-68,89,共4页
Computer Engineering and Applications
基金
博士基金
江西省科技项目"红壤地区石漠化遥感调查"
关键词
遥感
融合
LFF融合法
保持光谱
Remote sensing,Fusion,LFF Fusion Method,Spectr al preservation
