作　　者： ; ; ; ;

机构地区： 中山大学地理科学与规划学院

出　　处： 《北京测绘》 2013年第5期1-7,共7页

摘　　要： 高空间分辨率卫星影像的融合研究一直备受关注。本文以WorldView-2全色、多光谱影像为数据源,采用主成分分析、小波-主成分分析、高通滤波、HCS(Hypersherical color space)四种融合方法进行实验,并对融合效果做出定性及定量评价。研究结果表明,HCS法不仅显著提高影像空间细节表现力,而且有效地保持多光谱影像的光谱信息,其融合影像质量最高。 It has been paid much more attention to achieve high quality of fusion images with little spectral and spatial in- formation lessened. As a recently launched remote senor, WorlView2 can capture both image of high spatial resolution less than 0.5 meter in panchromatic band and multispectral images with 8 bands. In this study, selecting a part of Guang- zhou city as sample area, different methods such as principal component transformation, Wavelet-PC transformation, HPF and HCS will be used to fuse the images acquired from Worldview2. Fusion results will be further validated with qualitative analysis and quantitative evaluation. It is shown that the best quality of fusion image with more spectral and de- tailed information reserved can be derived from HCS method. It can provide a reference for selecting appreciate fusion method for the case of high spatial resolution images.

关 键 词： 遥感影像 高空间分辨率 融合

领　　域： [天文地球] [天文地球]