Infrared and visible images fusion based on contourlet-domain Hidden Markov Tree model

According to the fusion problem of infrared and visible images, the algorithm based on Contourlet-domain Hidden Markov Tree model (CHMT) is proposed in this paper. After the contourlet transform on the images, contourlet coefficients of the source images are trained to Contourlet-domain HMT model using the Expectation Maximization (EM) algorithm. Because the Contourlet-domain HMT model efficiently captures all dependencies across scales, space and directions through a tree structured dependence network, it can give more accurate description of images. Then a new fusion rule for the high frequency is built based on the window energy ratio, and weight average is adopted for low frequency. Experimental results show that the proposed algorithm provides more satisfying fusion results in terms of visual effect and objective evaluations, such as standard deviation, standard variance and clarity.