Multi-resolution Image Fusion Algorithm Based on Adaptive Fusion Rule

An adaptive fusion rule based on edge detection is designed for visible and infrared image fusion.Its main idea is that after detecting edges on high frequency sub-images of wavelet decomposition,for fusion of edge points,the rule based on maximum of absolute values with consistent verification is used,and for fusion of non-edge points,the rule based on region variance is adopted,in addition,for fusion of low frequency sub-image,low frequency sub-image of wavelet decomposition of visible or infrared image is used directly.From the different points,the proposed fusion rule is compared with other seven fusion rules.The experimental results show that the proposed fusion rule not only remains the edge feature of visible image,but also integrates the high gray value information of infrared image,furthermore,the contrast of target and background becomes better.