Atmospheric corrosion assessed from corrosion images using fuzzy Kolmogorov-Sinai entropy

Fuzzy Kolmogorov-Sinai (K-S) entropy was used to characterize the irregularity of the spatial information distribution in a corrosion image. The fuzzy K-S entropy was measured for horizontal and vertical orientations of corrosion images of steel samples exposed to a marine atmosphere over 191 days. The fuzzy K-S entropy for horizontal and vertical orientations decreased as the corrosion propagated. A relationship between the fuzzy K-S entropy and the weight loss in the steel sample can be presented with a linear mathematical expression. Fuzzy K-S entropy was found to be a semiquantitative and fast method to quantify atmospheric corrosion. (C) 2017 Elsevier Ltd. All rights reserved.