基于小波变换的IRS与TM遥感卫星影像融合

IRS and TM remote sensing image fusion based on wavelet transform

本研究在整体小波变换融合方法基础上 ,提出了基于对比度的局部特征选择加权小波变换影像融合方法 ,并以多光谱TM和高空间分辨率IRS 1C全色波段图像为例 ,与色彩空间变换HIS融合方法进行图像融合效果的比较分析试验。试验表明基于对比度的局部特征选择加权小波变换影像融合方法能在最大限度保持多光谱影像光谱信息的同时 ,增强了影像的纹理信息。

By image fusion, many kinds of remote sensing data can be synthesized, and supplementary mutual advantage can be realized by obtaining the exact information efficiently. This paper introduced the principles and characteristics of wavelet transform in fusion of remote sensing satellite image, and a new fusion method was then presented which is called partial selected weight mallat wavelet transform, based on contrast, after improving traditional wavelet transform. IRS 1C panchromatic and TM multi spectral image data were used in comparative experiment; the result was compared with HIS method in terms of image quality. It is showed that the proposed method maximizedly preserves the Texture information of the original IRS panchromatic data, as well as the spectral content of the TM data, and is better than other conventional methods.