構造物調査・診断(3)／振動

[V-672] Fundamental Study on Classification Efficiency for the Digital Photograph in the Inspection for COncrete Bridges by means of Machine Learning

〇Souta Sasawaki¹, Tomohiro Miki¹ (1.Kobe University)

Keywords:Concrete bridges, Maintenance, Classification of digital photos in the bridge inspection, Machine learning

This study presents a classification method of inspection photos of concrete bridges as an application of machine learning in the civil engineering field. The application to classify digital images by means of machine learning implemented using Python is developed. The classification of 10 to 100 images of the lines which were modeled concrete cracks was conducted. The results of the various situations of the images were analyzed and the accuracy and efficiency were discussed.