摘要
基于小波的多分辨率分析理论提出了一种新的遥感影像数据融合方法。它利用区域方差最大和一致性准则对不同尺度下的子带数据进行融合，采用加权运算对相应基带数据进行复合。文中给出了黑白航空影像与ＴＭ影像、ＳＡＲ影像的融合结果。通过与基于像素平均的融合方法比较，证明了本方法具有良好的鲁棒性和自适应能力。
This paper presents a new image fusion scheme based on multiresolution analysis of Wavelet theory for Remote Sensing imagery. An area-based maximum variance selection rule and a consistency detection are used for feature selection of subbands at different scale, and gray computation with weighting factor is accomplished for base-bands at the coarsest resolution. Experimental results including the fusion of black-white aerial imagery and Landsat(TM)imagery, Landsat and SAR images are presented in the paper. Compared with the pixel averaging method, the proposed scheme is more reliable and capable of adaptability.
出处
《中国图象图形学报（A辑）》
CSCD
1997年第2期103-107,共5页
Journal of Image and Graphics
基金
国防科工委资助
国家教委博士点基金
关键词
影像融合
多分辩率分析
遥感图像
Image fusion, Wavelet, Multiresolution analysis, Consistency detection
