Figure 4: Relative error average, after 100 runs, for the first model corresponding to Rice noise.