相似尺度图像融合算法

Image Fusion at Similar Scale

提出了一种新的融合思想,即图像应在相似尺度(si milar scale,SS)上进行融合·当融合低分辨率多光谱图像与高分辨率全色图像时,一般的方法没有考虑到插值的多光谱图像和高分辨率的全色图像的尺度不一致性·基于相似尺度的思想,图像融合算法如下·首先,使用“劋trous”离散小波变换分解高分辨率全色图像,使其低通分量与插值后的多光谱图像具有相似的尺度·然后,用加权多尺度基本形式(weighted mutlitscale fundamental form,WMFF)来融合它们得到合成的最低频带·最后,“劋trous”逆小波变换用来重建高分辨率的多光谱图像·与其他的基于小波变换的图像融合算法相比,基于相似尺度的融合方法取得了更好的融合结果·

An novel idea that image fusion should be performed at similar scale is proposed. Based on the idea “similar scale”, an image fusion algorithm is presented. At first, the “à trous” discrete wavelet transform is used to decompose the panchromatic image so that its approximation and the interpolated muhispectral image have the similar scale. Secondly, image fusion based on weighted muhiscale fundamental form （WMFF） is performed on them to obtain the synthesized approximation. Finally, the inverse “à trous” wavelet transform of the synthesized approximation and the detail of the panchromatic image reconstructs a high spatial resolution multispectral image. Compared with other wavelet-based techniques in the literature, the proposed image fusion algorithm has better performance.