3 files

Corrosion Condition State Semantic Segmentation Dataset

posted on 13.12.2021, 14:42 by Eric BianchiEric Bianchi, Matthew Hebdon

The data was collected from the Virginia Department of Transportation (VDOT) Bridge Inspection Reports. The data was semantically annotated following the corrosion condition state guidelines stated in the American Association of State Highway and Transportation Officials (AASHTO) and Bridge Inspector's Reference Manual (BIRM). There were four corrosion class categories: [good, fair, poor, severe]. The dataset consisted of 440 finely annotated images and was randomly split into 396 training images and 44 testing images. The images were resized to 512x512 for training and testing the DeeplabV3+ model. The original and resized images are included. After training with the DeeplabV3+ model (DOI: 10.7294/16628668), we were able to receive an F1 score of 86.67. More details of the training, the results, the dataset, and the code may be referenced in the journal article. The GitHub repository information may be found in the journal article.

If you are using the dataset in your work, please include both the journal article and the dataset citation.


National Science Foundation Grant No. IIS-1840044



University Libraries, Virginia Tech


English (US)


Virginia Tech