Technique for image fusion based on NSST domain INMF

Abstract A novel technique for image fusion based on non-subsampled shearlet transform (NSST) domain improved non-negative matrix factorization (INMF) is proposed. Firstly, NSST, which owns much lower computational complexity compared with other conventional typical multi-resolution tools, is adopted to perform the multi-scale and multi-directional decompositions of source images. Secondly, the traditional basic NMF model is updated to be an improved NMF (INMF). INMF is utilized to capture the marked characteristics in a series of sub-band components from the pure mathematical point of view and without destroying the two-dimensional structural information in the image. Thirdly, with INMF and the model of local directional contrast (LDC), the fused sub-images can be achieved. Finally, the final fused image can be obtained by using the inverse NSST. Experimental results demonstrate that the presented technique outperforms other typical NMF-based ones in both visual effect and objective evaluation criteria.

[1]  Ioannis Pitas,et al.  NMF, LNMF, and DNMF modeling of neural receptive fields involved in human facial expression perception , 2006, J. Vis. Commun. Image Represent..

[2]  Zhang Baohua,et al.  A multi-focus image fusion algorithm based on an improved dual-channel PCNN in NSCT domain , 2013 .

[3]  Bernt Schiele,et al.  Introducing a weighted non-negative matrix factorization for image classification , 2003, Pattern Recognit. Lett..

[4]  Shuyuan Yang,et al.  Fusion of multiparametric SAR images based on SW-nonsubsampled contourlet and PCNN , 2009, Signal Process..

[5]  Yi Chai,et al.  Multi-focus image fusion based on nonsubsampled contourlet transform and focused regions detection , 2013 .

[6]  Minh N. Do,et al.  The Nonsubsampled Contourlet Transform: Theory, Design, and Applications , 2006, IEEE Transactions on Image Processing.

[7]  Wang-Q Lim,et al.  Edge analysis and identification using the continuous shearlet transform , 2009 .

[8]  H. Sebastian Seung,et al.  Learning the parts of objects by non-negative matrix factorization , 1999, Nature.

[9]  屈小波 Xiaobo Qu,et al.  Image Fusion Algorithm Based on Spatial Frequency-Motivated Pulse Coupled Neural Networks in Nonsubsampled Contourlet Transform Domain , 2008 .

[10]  Yuanyuan Wang,et al.  Biological image fusion using a NSCT based variable-weight method , 2011, Inf. Fusion.

[11]  Xiao-Hui Yang,et al.  Fusion Algorithm for Remote Sensing Images Based on Nonsubsampled Contourlet Transform , 2008 .

[12]  Jianping Liu,et al.  Technique for image fusion based on nonsubsampled shearlet transform and improved pulse-coupled neural network , 2013 .

[13]  Demetrio Labate,et al.  Optimally Sparse Multidimensional Representation Using Shearlets , 2007, SIAM J. Math. Anal..

[14]  Reinhard Eckhorn,et al.  Neural mechanisms of scene segmentation: recordings from the visual cortex suggest basic circuits for linking field models , 1999, IEEE Trans. Neural Networks.

[15]  Yi Chai,et al.  Multifocus image fusion based on features contrast of multiscale products in nonsubsampled contourlet transform domain , 2012 .

[16]  Jingwen Yan,et al.  Image Fusion Algorithm Based on Spatia Frequency-Motivated Pulse Coupled Neural Networks in Nonsubsampled Contourlet Transform Domain: Image Fusion Algorithm Based on Spatia Frequency-Motivated Pulse Coupled Neural Networks in Nonsubsampled Contourlet Transform Domain , 2009 .

[17]  Paul L. Rosin,et al.  Robust Automatic Data Decomposition Using a Modified Sparse NMF , 2007, MIRAGE.

[18]  Jie Zhang,et al.  Technique for image fusion based on non-subsampled contourlet transform domain improved NMF , 2010, Science China Information Sciences.

[19]  Mei Yang,et al.  A novel algorithm of image fusion using shearlets , 2011 .

[20]  G. Easley,et al.  Sparse directional image representations using the discrete shearlet transform , 2008 .