小波变换保持光谱特型融合算法

A image fusion algorithm for spectrum respective based on wavelet

对遥感图像融合Brovey变换法存在颜色失真的现象,本文提出了一种小波比值融合法,该融合法首先对高几何分辨率的全色波段进行小波变换提取其低频分量,并对其进行三次卷积法重构得到与原分辨率相同的图像,而后将低分辨率多光谱图像与全色波段图像相乘,再除以重构后的全色波段图像,便得到融合图像。并从目视评价、GVI、统计评价证实了该小波比值融合法优于Brovey变换法,该比值融合法是一种能较好地保全低分辨率多光谱图像颜色的融合方法。

For remote sensing image fusion Brovey transform color distortion of the existence of the phenomenon,a wavelet ratio of integration,the integration of the first high geometric resolution panchromatic band wavelet transform from its low-frequency components,and its three volumes of Reconstruction of the original resolution by the same image,and will be low-resolution multi-spectral image and panchromatic band images multiply,and then divided by the Reconstruction after the panchromatic band image,they are fused image.And from the visual assessment,GVI,statistical evaluation confirmed that the ratio of fusion method is better than the wavelet transform Brovey,the integration of low-pass ratio is a better protection to low-resolution multi-spectral images of the color fusion.