改进的基于Curvelet变换的遥感图像融合方法

An Improved Image Fusion Method Based on Curvelet Transform

为了提高图像的视觉效果,提出了一种改进的基于Curvelet变换的图像融合方法.首先对全色图像进行增强,使图像的细节信息得到加强,再采用Curvelet变换与IHS变换结合的方法对增强后的图像进行融合。该算法既有效的保留了空间信息又抑制了光谱扭曲.结果表明,和常用的小波融合算法相比,笔者提出的融合方法在光谱信息的保持与空间细节信息的增强两个方面的综合性能得到提高,具有更好的效果.

An improved image fusion scheme based on Curvelet transform is proposed in this paper in order to enhance the visual effect.Firstly,the PAN image is enhanced in order to increase the edge information.Then,the enhanced image and MS image are fusioned based on Curvelet transform and IHS transform.This method can effectively retain the spatial information and reduce the spectral distortion.The results of fusion replace intensity component of multispectral image.Finally,the fusion image is obtained by taking the inverse IHS transform.The experiments prove the performance of the new scheme is better than those of typical WT fusion method in both of spatial and spectral domains.