基于二维EMD算法的多光谱全色图像融合

Panchromatic and Multi-spectral Image Fusion Based on 2D EMD Algorithm

为了获得图像的更多信息,全色图像和多光谱图像融合技术被广泛用于获取具有高空间分辨率的全色图像和多光谱图像的信息,以便更好地应用于军事、航空等领域。但目前,多光谱全色图像融合算法中存在色彩失真,波段受限等不足。为了获得图像信息,利用二维经验模式分解(EMD)的方法在处理非平稳信号等方面具有的明显优势,提出二维EMD的新颖图像融合算法。通过在二维经验模式分解(EMD)中创新性地引入开放角,算法将全色图像和多光谱图像分解成若干内蕴模式函数(IMF)和一个残差;然后,用全色图像的高频IMF替换多光谱图像的高频IMF;最后,通过重建混合多光谱图像的IMF获得融合图像。进行仿真,实验结果表明,改进的算法对图像融合是有效的,可为实际研究提供依据。

In order to obtain more information on the image,panchromatic and multispectral image fusion techniques are widely used to obtain high spatial resolution panchromatic and multispectral images,in order to better apply to the military,aviation and other fields.But panchromatic and multi-spectral image fusion algorithms have some shortcomings such as color distortion and limited band,to overcome the shortcomings,a novel image fusion algorithm is proposed based on two-dimensional empirical mode decomposition（EMD）,using obvious advantage of two-dimensional empirical mode decomposition（EMD） method in dealing with non-stationary signals.By the innovative introduction of open angle in two-dimensional EMD,this algorithm based on EMD is to decompose the panchromatic and multi-spectral images into their IMFs（Intrinsic Mode Functions） and a residual.Then,the high frequency IMF of multi-spectral image is replaced by high frequency IMF of panchromatic image.Finally,the image fusion is performed by reconstructing the mixed IMFs.The experimental results indicate that the proposed method is able to product a smaller distortion in fusion image and is effective.