基于不同算法的遥感影像融合分析  

Analysis of Remote Sensing Image Fusion Based on Different Algorithms

作　　者：王广杰[1,2] 周介铭[2] 杨存建[2] 赵景峰[2] 

机构地区：[1]成都理工大学地球科学学院,四川成都610059 [2]四川师范大学西南土地资源评价与监测教育部重点实验室,四川成都610066

出　　处：《四川师范大学学报（自然科学版）》2011年第2期255-259,共5页Journal of Sichuan Normal University（Natural Science）

基　　金：国家"863"计划(2009AA12Z140);国家自然科学基金(40771144和40575035);四川省教育厅重点科研基金(09ZA088)资助项目

摘　　要：遥感影像融合使影像具有高空间分辨率和丰富的光谱信息,遥感影像融合算法是近年来遥感领域内的研究热点,而针对不同的遥感影像,有关学者提出了很多融合算法,但是融合影像的光谱信息保真度和空间信息融入度却一直是遥感融合效果研究的焦点问题,因此有必要对融合方法进行评价.以ETM+影像为例,通过ENVI软件作为遥感影像处理平台,选择均值、标准差、偏差指数、相关系数等评价指标,进行PCA变换、Multiplicative变换、Brovey变换和Wavelet变换等4种全色与多光谱遥感影像融合算法试验,并对融合试验结果进行定量评价,探讨ETM+影像融合的最佳方法.结果表明,最佳的遥感影像融合算法是Wavelet变换融合,Multiplicative变换融合要优于Brovey变换融合和PCA变换融合.这一研究结果为充分利用全色高分辨率和多光谱特性,挖掘的数据潜力奠定了基础.Remote sensing（RS） image fusion makes the image has high resolution and abundant spectrum information,and the fusion algorithm is a study hot spot in RS field in recent years.Though many fusion algorithms for different RS images have been put forward,the fidelity of spectrum information and the fusion degree of spatial information are still focus issues in fusion algorithm study.Therefore,it is necessary for researchers to evaluate the effect of different fusion algorithms.In this paper,ETM＋ image is selected based on ENVI software platform.The fusion results of panchromate and multi-spectra are quantitatively evaluated and the best fusion algorithm is discussed by different fusion algorithms,such as PCA,Multiplicative,Brovey and Wavelet transformations applied for different index,such as average value,standard deviation,deviation index,and relative coefficient.The result shows that Wavelet transformations is the best fusion algorithm,and Multiplicative transformations is better than Brovey and PCA.This result has laid a foundation to make full use of the panchromatic high resolution and multi-spectral charactristics,and to mine data potential.

关 键 词：ETM+影像 数据融合 算法 全色与多光谱 定量评价 

分 类 号：P237.3[天文地球—摄影测量与遥感]