Feature extraction and quantitative analysis of steel corrosion in reinforced concrete components based on XCT scanning and deep learning model

Accurate quantification and analysis of steel corrosion is crucial for reliability assessment studies of in-service reinforced concrete structures . However, the pixel-level cross-sectional data provided by X-ray computed tomography (XCT) proves difficult to quantify, especially for the amorphous corrosion products filled in mortar, due to the absence of robust feature extraction methods. In this study, multiple deep learning models were trained to automatically identify corrosion products and calculate corrosion-related parameters from a large number of XCT images. The database comprised XCT images obtained from a RC component subjected to chloride-rich environment for four years. The results indicate that deep learning models can segment different regions of XCT images with high accuracy. Among the models, the K-Net model performed the best on this dataset, achieving an accuracy of 94.60 %, and a mean Precision (mPrecision) of 88.21 %. This advance makes it possible to automatically extract parameters that characterize steel corrosion and to assess the damage to RC structures caused by corrosion. • Deep learning models and XCT images enable automated quantitative analysis of internal corrosion behavior in concrete. • This study constructed a database using XCT images of RC components that underwent 4-year wet-dry cycles in chloride-rich environment. • The best-performing model on this dataset was the K-Net model, which achieved an accuracy of 94.60 % on the test dataset. • This method is automatic, accurate, and fast, and is suitable for analyzing a large number of XCT images obtained from RC components. • The study explored the relationships and mechanisms between parameters related to steel corrosion.