Image analysis of atmospheric corrosion exposure of zinc

A computational approach using wavelet transformation and image processing methods to analyze and quantify the extent of corrosion damage from the images of zinc exposure samples is described and discussed. This method contains four steps. First, a scanner was used to acquire corrosion morphology images of non-ferrous metal exposed to the atmospheric environment and the noise was then eliminated from the acquired images; in the second step, wavelet transformation was applied to decompose the improved images; then, the energies of the sub-images were extracted as feature information; finally, the relationship between the features and corrosion weight loss of samples was evaluated by regression analysis. By using the equation given by the regression analysis, the zinc atmospheric corrosion weight loss can be calculated.