Thesis - Open Access
Master of Science (MS)
Electrical Engineering and Computer Science
Computer Vision, Deep Learning, Machine Learning
Damages assessment of bridges is important to derive immediate response after severe events to decide serviceability. Especially, past earthquakes have proven the vulnerability of bridges with insufficient detailing. Due to lack of a national and unified post-earthquake inspection procedure for bridges, conventional damage assessments are performed by sending professional personnel to the onsite, detecting visually and measuring the damage state. To get accurate and fast damage result of bridge condition is important to save not only lives but also costs.
There have been studies using image processing techniques to assess damage of bridge column without sending individual to onsite. Convolutional neural networks (CNNs) have shown state-of-art results in object detection and image classification tasks. This study proposed cascaded deep learning network for post-earthquake bridge serviceability assessment. Major target deficiency components (crack, spalled area, transverse bar, and longitudinal bar) were used to determine the proposed damage states to assess serviceability of bridge. Cascaded network is composed by Mask R-CNN and MobileNet v2 which have been proved as powerful network for each instance segmentation and image classification.
In this study, proposed network successfully detected target deficiency components and measured each damage state by following 5 stages. Column area is detected as first step, and counting exposed bars, finding maximum distance in spalled region within column area are followed to decide damage state. To determine deficiency of crack in bridge column, crack patch classification module is attached in proposed network. Counting diagonal and horizontal cracks with angle measurement are used to analyze type of cracks.
Number of Pages
South Dakota State University
In Copyright - Non-Commercial Use Permitted
Jang, Youjeong, "Cascaded Deep Learning Network for Postearthquake Bridge Serviceability Assessment" (2021). Electronic Theses and Dissertations. 5707.