A Non-Intrusive System to Classify the Severity of Damages Caused by Internal Corrosion Using the Potential Drop Technique and Electrical Image Mapping

Abstract

This work presents a non-intrusive method to obtain information about damages caused by internal corrosion in a stainless-steel plate and classify them according to their severity. The Potential Drop technique provides an electric potential gradient map, which is analyzed by the application of image processing techniques, such as morphological analysis and segmentation. Some corrosion forms can be detected by this method, like cracks and pitting corrosion; the last one is discussed in this paper. Finite Element Modeling simulations were performed to get examples of defective plates (with two classes of damages) The image processing in the simulations acts as a feature extractor that feeds a binary classifier based on Logistic Regression, which accuracy was 99.24%.