Complex wavelet domain image fusion based on fractional lower order moments

We describe an image fusion algorithm for data exhibiting heavy tails with no convergent second- or higher-order moments. Our developments rely on recent results showing that wavelet decomposition coefficients of images are best modeled by alpha-stable distributions, a family of heavy-tailed densities. Thus, in the multiscale wavelet domain we develop a novel fusion rule based on fractional lower order moments (FLOM's). Simulation results show that our method achieves better performance in comparison with previously proposed pixel-level fusion approaches.