ABSTRACT
In recent years, the structural integrity of social infrastructures has become a concern as they age, and how to efficiently and accurately assess their structural integrity is important in developing maintenance management plans. In this study, we aim to save time and labor by automating the evaluation of corrosion progress using machine learning. Specifically, we will investigate whether it is possible to detect corrosion based on deep learning using images of monorail bridges taken from an on-board camera, and then evaluate the progress of corrosion using images taken at the different inspection time.
