A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Automated corrosion detection using deep learning and computer vision
https://orcid.org/http://orcid.org/0000-0001-6434-3828
https://orcid.org/http://orcid.org/0000-0001-8123-3938
Over the past few decades, there has been an increase in the number of aging civil structures worldwide, most of which are made of concrete. Concrete may lose strength as a result of continuous loading and environmental factors. It has been challenging to detect corrosion damage in industrial and civil constructions, and the existing structures must be maintained manually, which has proven to be subjective and unreliable. Thus, it is essential to have a system for detecting defects in concrete structures, such as seeing an automatic corrosion product which could early show the warning of damages to avoid disasters caused by structural failures. This research used three versions of the You Only Look Once (YOLO) object detection technique to identify concrete corrosion from real-world images. The training was conducted using concrete corrosion images for YOLOv3, YOLOv5s, and YOLOv7 models. The performance of YOLOv3, YOLOv5s, and YOLOv7 was compared using evaluation metrics such as accuracy, F1 score, recall, and mean average precision (mAP). The results showed that YOLOv3, YOLOv5s, and YOLOv7 had F1 scores of 0.85, 0.83, and 0.80, respectively. Among all these models, YOLOv5s achieved the highest mAP@0.5 values of 0.88 which outperforms other state-of-the-art detectors.
Automated corrosion detection using deep learning and computer vision
https://orcid.org/http://orcid.org/0000-0001-6434-3828
https://orcid.org/http://orcid.org/0000-0001-8123-3938
Over the past few decades, there has been an increase in the number of aging civil structures worldwide, most of which are made of concrete. Concrete may lose strength as a result of continuous loading and environmental factors. It has been challenging to detect corrosion damage in industrial and civil constructions, and the existing structures must be maintained manually, which has proven to be subjective and unreliable. Thus, it is essential to have a system for detecting defects in concrete structures, such as seeing an automatic corrosion product which could early show the warning of damages to avoid disasters caused by structural failures. This research used three versions of the You Only Look Once (YOLO) object detection technique to identify concrete corrosion from real-world images. The training was conducted using concrete corrosion images for YOLOv3, YOLOv5s, and YOLOv7 models. The performance of YOLOv3, YOLOv5s, and YOLOv7 was compared using evaluation metrics such as accuracy, F1 score, recall, and mean average precision (mAP). The results showed that YOLOv3, YOLOv5s, and YOLOv7 had F1 scores of 0.85, 0.83, and 0.80, respectively. Among all these models, YOLOv5s achieved the highest mAP@0.5 values of 0.88 which outperforms other state-of-the-art detectors.
Automated corrosion detection using deep learning and computer vision
https://orcid.org/http://orcid.org/0000-0001-6434-3828
https://orcid.org/http://orcid.org/0000-0001-8123-3938
Over the past few decades, there has been an increase in the number of aging civil structures worldwide, most of which are made of concrete. Concrete may lose strength as a result of continuous loading and environmental factors. It has been challenging to detect corrosion damage in industrial and civil constructions, and the existing structures must be maintained manually, which has proven to be subjective and unreliable. Thus, it is essential to have a system for detecting defects in concrete structures, such as seeing an automatic corrosion product which could early show the warning of damages to avoid disasters caused by structural failures. This research used three versions of the You Only Look Once (YOLO) object detection technique to identify concrete corrosion from real-world images. The training was conducted using concrete corrosion images for YOLOv3, YOLOv5s, and YOLOv7 models. The performance of YOLOv3, YOLOv5s, and YOLOv7 was compared using evaluation metrics such as accuracy, F1 score, recall, and mean average precision (mAP). The results showed that YOLOv3, YOLOv5s, and YOLOv7 had F1 scores of 0.85, 0.83, and 0.80, respectively. Among all these models, YOLOv5s achieved the highest mAP@0.5 values of 0.88 which outperforms other state-of-the-art detectors.
Automated corrosion detection using deep learning and computer vision
Asian J Civ Eng
Nabizadeh, Elham (author) / Parghi, Anant (author)
Asian Journal of Civil Engineering ; 24 ; 2911-2923
2023-12-01
13 pages
Article (Journal)
Electronic Resource
English
Corrosion detection with computer vision and deep learning
| TIBKAT | 2022
Automated Region-Based Vehicle Conflict Detection Using Computer Vision Techniques
| British Library Online Contents | 2015
Structural Damage Detection of Steel Corrugated Panels Using Computer Vision and Deep Learning
| Springer Verlag | 2024