Deep learning based automated segmentation of air-void system in hardened concrete surface using three dimensional reconstructed images. (21st March 2022)

• Record Type: Journal Article
• Authors: Tao, Jueqiang; Gong, Haitao; Wang, Feng; Luo, Xiaohua; Qiu, Xin; Liu, Jinli
• Is Part Of: Construction & building materials. Volume 324(2022)
• Publication Date: 2022-03-21
• Subjects: 3D reconstruction -- DCNN -- Semantic segmentation -- Air voids -- Hardened concrete; Building materials -- Periodicals; 624.18
• DOI: 10.1016/j.conbuildmat.2022.126717
• Languages: English
• ISSNs: 0950-0618

Abstract

Highlights: Automated segmentation of air voids from hardened concrete surface. Three-dimensional image reconstruction method. Deep learning strategies and process. Abstract: The automated air-void detection methods specified in the ASTM C457 require the aid of contrast enhancement which is time consuming and labor intensive. This study investigated the utilization of three-dimensional (3D) reconstruction and Deep Convolution Neural Network (DCNN) methods to detect the air voids in hardened concrete surfaces without the use of contrast enhancement. The experimental results showed that the DCNN could accurately distinguish air voids from hardened concrete images with the detection accuracy of over 0.9 in only less than a minute. The accuracy rates for air content, specific surface, and spacing factor were 0.92, 0.91, and 0.89, respectively.