Color Image Fusion Algorithm Using the Contourlet Transform

LI Guang-xin,WANG Ke

DOI: https://doi.org/10.3321/j.issn:0372-2112.2007.01.023

2007-01-01

Abstract:With the particular research on thermal and visual images,a color image fusion algorithm using the contourlet transform is presented.Firstly,through the IHS(Intensity-Hue-Saturation) transform,the color visual image is converted from RGB color space to IHS space.Next,with the contourlet transform and weighted average fusion rule,the intensity component and thermal image are merged into a grayscale image,which is then linearly stretched to have the same mean and variance as the intensity component.Finally,the stretched grayscale fused image replaces the original intensity component,and the final RGB color fused image is achieved by the inverse IHS transform with the H,S and replacement component.On the one hand,with the proposed scheme,the contourlet transform as a new mathematical tool is introduced to image fusion area.On the other hand,the algorithm provided a new color image fusion strategy of thermal and visual images.The experimental results show that,with the proposed color fusion method,the fused image produced by the contourlet transform is of better quality than that obtained through the wavelet transform.Moreover,the color fusion approach obviously improves fusion performance over the traditional IHS transform fusion method.