IMAGE PROCESSING TECHNIQUES APPLIED FOR PITTING CORROSION ANALYSIS

Abstract

In order to study the behavior of the early stage of pitting corrosion, an image analysis based on discrete wavelet packet transform and fractals was used. Image feature parameters were extracted and analyzed to characterize the pitting corrosion development with test time. It was found that the feature parameters: Shannon entropy, energy, fractal dimension and intercept increased with the test time. Therefore the image processing techniques were promising and effective tools to analyze and detect the pitting corrosion.