Figure 6: Relative error average, after 100 runs, for the second model corresponding to Rice noise.