Image fusion based on nonsubsampled contourlet transform

摘要

In this work we have presented an image fusion method based on nonsubsampled contourlet transform (NSCT). Nonsubsampled contourlet transform is a multiscale, multidirectional transform with shift-invariance feature. Shift-invariance is a desirable property for image fusion and many other applications of image processing because it leads to several advantages such as removal of pseudo-Gibbs phenomenon, better frequency selectivity, regularity, improved temporal stability and consistency. Here the input images are decomposed by nonsubsampled contourlet transform and then different fusion rules are applied for low and high frequency NSCT coefficients. Finally the fused image is obtained by the inverse transform. To validate the effectiveness of the proposed method, visual and quantitative analysis are also performed. From the obtained results it is clear that the proposed method outperforms other contourlet and wavelet based methods.