Bio-medical image fusion based on phase-congruency and guided filter

In this paper, we propose an approach for fusion of the bio-medical images based on phase-congruency (PC) and guided filter. In the first step, we decompose the bio-medical source images into low and high frequency sub-bands by using non-subsampled contourlet transform (NSCT). The low frequency sub-bands are processed further to extract image features with the use of PC. Literature shows that the PC is a powerful mathematical operation for image analysis which is the biologically inspired operation towards detection of the image features that are invariant to image illumination and contrast. The high frequency details are extracted from the high frequency NSCT sub-bands by using a guided filter which is proven to be very effective in order to preserve the edge details with smooth regions. In the last, inverse NSCT is applied on the fused NSCT sub-bands to obtain the final fused image in spatial domain. The efficacy of the proposed approach is analyzed by conducting different experiments on real bio-medical images acquired by different multi-modal medical sensors. The fusion results are compared with the other state of the art approaches which show that the proposed approach performs better in terms of visual and quantitative evaluations.