Development of a Deep Learning System with Small Amount of Data for Bridges Health Monitoring

Abstract

In this paper, we show the deep learning system for the health monitoring system of bridges. It checks two factors in the bridges, clacks and corrosions (or rusts). We try to make the deep learning system with small amount of data (about thousands of data). We have constructed several CNN models using fine tuning. We are able to make the accuracy CNN model to employ fine tuning with small training data. We achieved 86.75% average accuracy in Xception fine tuning model for clacks and 83% accuracy for corrosions.  Furthermore, we have embedded the CNN models into Jetson Nano which is one of the embedded boards.