| Title |
Surface corrosion detection for ferrous-metal parts: application of artificial intelligence, python and microscopic images |
| Authors |
Wdowik, Roman ; Bełzo, Artur ; Bendikiene, Regita |
| DOI |
10.5755/j02.ms.41377 |
| Full Text |
|
| Is Part of |
Materials science = Medžiagotyra.. Kaunas : KTU Technologija. 2025, vol. 31, no. 3, p. 431-438.. ISSN 1392-1320. eISSN 2029-7289 |
| Keywords [eng] |
damage identification ; corrosion ; surface ; steel ; artificial intelligence |
| Abstract [eng] |
This paper presents a novel method for the identification of surface damage, in particular corrosion, in ferrous metals based on generative artificial intelligence (GenAI), showing how to automate damage identification and corrosion recognition. The methodology involved using optical microscopy to capture electrochemical corrosion patterns, followed by image preprocessing and classification using AI algorithms implemented in Python. High-quality microscopic images have been recorded, based on selected ferrous metals. Python code lines were generated using ChatGPTTM based on queries created by the authors, and this method was applied to the corrosion analysis. Quantitative evaluation confirmed Python code parameters-dependent detection accuracy and repeatability, demonstrating the robustness of the proposed technique. The results were discussed in terms of possible industrial applications. In addition, the limitations of the results obtained, which sometimes fall short of the claims inspector's expectations, were discussed. Compared to traditional corrosion detection methods such as visual inspection and non-destructive testing, AI-based methods are a faster and more cost-effective solution that can process large volumes of images in real time and produce consistent results. Further research directions are also suggested, including the analysis of other types of damage and improving the accuracy of the model. In addition to technical efficiencies, the broader impact of these studies is that they can contribute to predictive maintenance, reduce downtime and improve safety in industries with high ferrous metal use. |
| Published |
Kaunas : KTU Technologija |
| Type |
Journal article |
| Language |
English |
| Publication date |
2025 |
| CC license |
|
