Image fusion based on object region detection and Non-Subsampled Contourlet Transform

Abstract This paper presents a new fusion algorithm for infrared (IR) and visible light (ViS) images by combining object region detection with the Non-Subsampled Contourlet Transform (NSCT). First, the saliency map for the IR image is acquired with saliency detection. Second, the object region in the IR image is extracted by introducing a free regions removal method. Third, source images are decomposed via NSCT and different fusion rules for low frequency sub-bands and high frequency sub-bands are employed. Then, the primary fused image is generated by the inverse NSCT. Finally, integrating the primary fused image with the object region, the final fused image is obtained. By conducting experiments, we compare our method to others using several metrics and results show that the proposed method can improve the quality of the fused image.

[1]  Mingliang Xu,et al.  High quality multi-spectral and panchromatic image fusion technologies based on Curvelet transform , 2015, Neurocomputing.

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

[3]  Luciano Alparone,et al.  Remote sensing image fusion using the curvelet transform , 2007, Inf. Fusion.

[4]  Eduardo A. B. da Silva,et al.  Multiscale Image Fusion Using the Undecimated Wavelet Transform With Spectral Factorization and Nonorthogonal Filter Banks , 2013, IEEE Transactions on Image Processing.

[5]  Hongxun Yao,et al.  Image Evaluation Factors , 2005, ICIAR.

[6]  Paul M. de Zeeuw,et al.  Fast saliency-aware multi-modality image fusion , 2013, Neurocomputing.

[7]  Liu Cao,et al.  Multi-Focus Image Fusion Based on Spatial Frequency in Discrete Cosine Transform Domain , 2015, IEEE Signal Processing Letters.

[8]  Jürgen Jahns,et al.  Efficient Hadamard transformation of large images , 1983 .

[9]  Zheng Liu,et al.  Image fusion by using steerable pyramid , 2001, Pattern Recognit. Lett..

[10]  Wenzhan Dai,et al.  Fusion algorithm of infrared and visible images based on local energy using NSCT , 2012, Proceedings of the 10th World Congress on Intelligent Control and Automation.

[11]  Qiang Zhang,et al.  Multifocus image fusion using the nonsubsampled contourlet transform , 2009, Signal Process..

[12]  W. Kong,et al.  Fusion technique for grey-scale visible light and infrared images based on non-subsampled contourlet transform and intensity-hue-saturation transform , 2011 .

[13]  R. P. Desale,et al.  Study and analysis of PCA, DCT & DWT based image fusion techniques , 2013, 2013 International Conference on Signal Processing , Image Processing & Pattern Recognition.

[14]  S. Süsstrunk,et al.  Frequency-tuned salient region detection , 2009, CVPR 2009.

[15]  K. Muneeswaran,et al.  Region-based saliency detection , 2014, IET Image Process..

[16]  Zheng Liu,et al.  Directive Contrast Based Multimodal Medical Image Fusion in NSCT Domain , 2013, IEEE Transactions on Multimedia.

[17]  Yaonan Wang,et al.  Combination of images with diverse focuses using the spatial frequency , 2001, Inf. Fusion.