Muon tomography is a technique that aims to reconstruct the internal composition of an unknown volume by analyzing the trajectories of muons which traverse it. We propose to extend the reference method of analysis (Schultz et al., 2007) by using a mixture of two Gaussian distributions to better capture the heavy tails observed in the distribution of the muon scattering. The proposed model is fitted using stochastic EM. An application aimed at evaluating the wear level in the inner walls of an insulating tube is given.