To improve the efficiency and reduce the labour cost of the renovation process, this study presents a lightweight Convolutional Neural Network (CNN)-based architecture to extract crack-like features, such as cracks and joints. Moreover, Transfer Learning (TF) method was used to save training time while offering comparable prediction results. For three different objectives: 1) Detection of the concrete cracks; 2) Detection of natural stone cracks; 3) Differentiation between joints and cracks in natural stone; We built a natural stone dataset with joints and cracks information as complementary for the concrete benchmark dataset. As the results shown, our model is demonstrated as an effective tool for the industry use.
CITATION STYLE
Yang, J., Lin, F., Xiang, Y., Katranuschkov, P., & Scherer, R. J. (2021). Fast Crack Detection Using Convolutional Neural Network. In EG-ICE 2021 Workshop on Intelligent Computing in Engineering, Proceedings (pp. 540–549). Technische Universitat Berlin. https://doi.org/10.17485/ijst/v14i10.1245
Mendeley helps you to discover research relevant for your work.