Estimation of errors in determining corrosion grain sizes by analysis of diffuse light reflection signal

The approaches to estimation of lower boundary of the inverse problem solution error concerning the sizing the corrosion micro defects inside the submillimeter corrosion spots has been proposed. It is assumed that pointed error depends on random location of the corrosion spots. The method based on comparison of two estimations of light diffusion reflectance sensor’s signal discrepancy. The first estimation is based on the standard deviation for the discrepancy caused by randomly located corrosion spots. The second one is based on corrosion grains’ size deviation. Also, it is found that the discrepancy based on deviations of signal peaks positions provides more stable solution for the corrosion micro defects sizes.