In this paper, the area of mill-cut damage in an aluminium plate was identified with a 2D Wavelet Transform technique using 11 numerically simulated vibrational mode shape signals. These vibrational mode shapes were corrupted with various levels of artificial noise of a uniform distribution and reduced by integer values to simulate the performance of damage detection algorithm in real life situations with different sensor densities. The damage was assessed through the calculation of damage indices (DI) over the entire area of the plate and subsequent standardization, yielding standardized damage indices which are based on statistical hypothesis approach. These damage indices were summed over all vibrational modes considered in this study. The largest peak in the damage index profile revealed the location of damage. The confidence of damage identification given in percentage was estimated with a parameter called damage estimate reliability (DER). Results suggested that only an isotropic Pet Hat wavelet yielded satisfactory damage identification. Analysis of DER vs wavelet scale was carried out and results indicate that DER values rapidly decreased as scale increased, thus only 1st scale with a corresponding DER of about 98 % was adopted for further studies. The influence of various sensor densities on the damage estimate reliability was studied.