A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Automatic Detection of Brick Pavement Defects Using Convolutional Neural Network
Discovering the pavement defect is critical for maintaining the pavement health condition. However, a majority of pavement defect inspections are manual processes, which are time-consuming and labor-intensive. This process heavily relies on engineers’ experiences, and thus the judgments are prone to human error. To solve the limitations, this research achieves detection on brick pavement defects using a convolutional neural network. The Mask R-CNN is trained and validated on a subset of 600 and 100 annotated 512 × 512-pixel labeled images for segmentation, respectively. The Mask R-CNN to detect brick pavement defects: spalling, efflorescence, and hole obtains high precisions of 94.62%, 90.79%, and 93.75%, and high recalls of 84.06%, 73.67%, and 71.43%, respectively. The results demonstrate that the proposed method has a good performance and can achieve the detection of different brick pavement defects at the pixel level. This research not only benefits the extensive application of the method, but also can help engineers improve productivity for assessing the pavement conditions in operation and maintenance.
Automatic Detection of Brick Pavement Defects Using Convolutional Neural Network
Discovering the pavement defect is critical for maintaining the pavement health condition. However, a majority of pavement defect inspections are manual processes, which are time-consuming and labor-intensive. This process heavily relies on engineers’ experiences, and thus the judgments are prone to human error. To solve the limitations, this research achieves detection on brick pavement defects using a convolutional neural network. The Mask R-CNN is trained and validated on a subset of 600 and 100 annotated 512 × 512-pixel labeled images for segmentation, respectively. The Mask R-CNN to detect brick pavement defects: spalling, efflorescence, and hole obtains high precisions of 94.62%, 90.79%, and 93.75%, and high recalls of 84.06%, 73.67%, and 71.43%, respectively. The results demonstrate that the proposed method has a good performance and can achieve the detection of different brick pavement defects at the pixel level. This research not only benefits the extensive application of the method, but also can help engineers improve productivity for assessing the pavement conditions in operation and maintenance.
Automatic Detection of Brick Pavement Defects Using Convolutional Neural Network
Ji, Ankang (author) / Xue, Xiaolong (author) / Dou, Yudan (author) / Wang, Yuna (author)
International Conference on Construction and Real Estate Management 2021 ; 2021 ; Beijing, China
ICCREM 2021 ; 255-263
2021-12-09
Conference paper
Electronic Resource
English
Automatic classification of pavement crack using deep convolutional neural network
| Taylor & Francis Verlag | 2020