基于小波变换的MODIS与ETM数据融合研究  

Fusion of MODIS and ETM Images by Using Wavelet Transform

作　　者：余钧辉[1] 张万昌[1] 乐通潮[1] 

机构地区：[1]南京大学国际地球系统科学研究所,南京210093

出　　处：《遥感信息》2004年第4期39-42,F003,共5页Remote Sensing Information

基　　金：国家重点基础研究发展规划项目 ( 2 0 0 1CB3 0 940 4);海外青年学者合作研究基金 ( 4 0 12 80 0 1);教育部科学技术重点项目 ( 2 0 0 1)

摘　　要：遥感影像融合方法多样 ,但针对空间分辨率相差十倍、甚至十几倍的不同数据源影像进行融合的研究很有限 ,有效算法也较少。MODIS影像高光谱数据具有 36个相互配准的光谱波段信息[1～ 2 ] ,然而其 0 .2 5km～ 1km的低空间分辨率 ,却限制了其应用潜力。本文基于小波变换的算法思想提出了一种MODIS与Landsat +ETM(空间分辨率 30m)数据融合的方法 ,能够有效的将MODIS的光谱信息和ETM的空间几何信息结合起来 ,并在此基础上分析了地形阴影对融合的影响 ,为MODIS数据用于制作较大比例尺的土地利用现状图等提供了可能。Many methodologies can be found in literature on data fusion of various remotely sensed images, but few of them are available in data fusion of the images whose spatial resolutions are differed more than or equal to ten times. MODIS hyperspectral data has rich spectral information with 36 bands, unfortunately, is limited to many uses because of its low spatial resolution (0.25Km^1Km). The paper presents a methodology based on wavelet transform algorithm to fuse MODIS and ETM image (spatial resolution: 30m) data for improving the MODIS spatial resolution while keeping the advantage of MODIS hyperspectral information as many as possible. The experimental study suggested that this methodology can effectively achieve the goal as wanted, and the effects of topographical shadow to the fusion are analyzed with the example data. The method offers a possibility to use MODIS data in high resolution land use/land cover mapping.

关 键 词：小波变换 高光谱信息 MODIS landsat+ETM 融合 

分 类 号：P237.3[天文地球—摄影测量与遥感]