A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Real-Time Detection of Personal Protective Equipment Violations for Construction Workers Using Semisupervised Learning and Video Clips
https://orcid.org/0000-0003-4701-4533
The collection of personal protective equipment (PPE) violation data is crucial for assessing safety risks in behavior-based safety (BBS) management on construction sites. Owing to the limitations of manual inspection methods, many studies have employed computer vision–based methods for PPE violation detection. However, limitations exist in implementing this in actual construction projects due to the costs associated with the acquisition and labeling of a large number of images, and the accuracy and efficiency of recording PPE violation data. Therefore, this study introduces semisupervised object detection (SS-OD) and data augmentation for training PPE detection models to reduce the use of labeled data without reducing the performance, and proposes a framework for recording the PPE violation data by extracting PPE violation videos from real-time surveillance footage by considering the role of the worker and their position, thereby enhancing practical construction management and serving as input for BBS-based safety risk assessments on construction sites. The results show that (1) the labeled data demand for training PPE detection models can be reduced through semisupervised learning, image augmentation, and transfer learning, without reducing the performance of PPE detection, thereby reducing the cost of application on actual construction sites; (2) SS-OD methods are better equipped to handle changes in construction scenarios by making full use of unlabeled data, and thus are suitable for construction scenarios; and (3) recording PPE violation data using the video clip method in a real construction project achieved an average precision of 91.76% and an average score of 92.79%. Using the PPE detection model trained with SS-OD effectively records PPE violation data for BBS-based safety risk assessment. This study significantly enhances the efficiency of real construction site PPE violation inspections and provides a valuable method for the automated and real-time collection of violation data in BBS-based management.
Computer vision–based personal protective equipment violation detection methods from past studies have been difficult to implement widely in real projects due to cost and effectiveness limitations. This study effectively improves upon these limitations using semisupervised learning and video clipping methods. Experimental results demonstrate that the object detection model trained with the semisupervised learning method not only reduces the need for labeled data in personal protective equipment detection model training but also exhibits improved generalization capabilities in complex and dynamic construction scenarios. Using the personal protective equipment violation video clipping method proposed in this study allows for efficient editing of violation video clips, facilitating rapid identification and documentation of construction workers’ personal protective equipment violation behaviors. This enhances the safety culture at construction sites and aids construction managers in developing safety management measures and educational plans. The real-time-collected personal protective equipment violation data can be utilized further as input for a behavior-based safety risk assessment model, enhancing the automation of unsafe behavior data collection.
Real-Time Detection of Personal Protective Equipment Violations for Construction Workers Using Semisupervised Learning and Video Clips
https://orcid.org/0000-0003-4701-4533
The collection of personal protective equipment (PPE) violation data is crucial for assessing safety risks in behavior-based safety (BBS) management on construction sites. Owing to the limitations of manual inspection methods, many studies have employed computer vision–based methods for PPE violation detection. However, limitations exist in implementing this in actual construction projects due to the costs associated with the acquisition and labeling of a large number of images, and the accuracy and efficiency of recording PPE violation data. Therefore, this study introduces semisupervised object detection (SS-OD) and data augmentation for training PPE detection models to reduce the use of labeled data without reducing the performance, and proposes a framework for recording the PPE violation data by extracting PPE violation videos from real-time surveillance footage by considering the role of the worker and their position, thereby enhancing practical construction management and serving as input for BBS-based safety risk assessments on construction sites. The results show that (1) the labeled data demand for training PPE detection models can be reduced through semisupervised learning, image augmentation, and transfer learning, without reducing the performance of PPE detection, thereby reducing the cost of application on actual construction sites; (2) SS-OD methods are better equipped to handle changes in construction scenarios by making full use of unlabeled data, and thus are suitable for construction scenarios; and (3) recording PPE violation data using the video clip method in a real construction project achieved an average precision of 91.76% and an average score of 92.79%. Using the PPE detection model trained with SS-OD effectively records PPE violation data for BBS-based safety risk assessment. This study significantly enhances the efficiency of real construction site PPE violation inspections and provides a valuable method for the automated and real-time collection of violation data in BBS-based management.
Computer vision–based personal protective equipment violation detection methods from past studies have been difficult to implement widely in real projects due to cost and effectiveness limitations. This study effectively improves upon these limitations using semisupervised learning and video clipping methods. Experimental results demonstrate that the object detection model trained with the semisupervised learning method not only reduces the need for labeled data in personal protective equipment detection model training but also exhibits improved generalization capabilities in complex and dynamic construction scenarios. Using the personal protective equipment violation video clipping method proposed in this study allows for efficient editing of violation video clips, facilitating rapid identification and documentation of construction workers’ personal protective equipment violation behaviors. This enhances the safety culture at construction sites and aids construction managers in developing safety management measures and educational plans. The real-time-collected personal protective equipment violation data can be utilized further as input for a behavior-based safety risk assessment model, enhancing the automation of unsafe behavior data collection.
Real-Time Detection of Personal Protective Equipment Violations for Construction Workers Using Semisupervised Learning and Video Clips
https://orcid.org/0000-0003-4701-4533
The collection of personal protective equipment (PPE) violation data is crucial for assessing safety risks in behavior-based safety (BBS) management on construction sites. Owing to the limitations of manual inspection methods, many studies have employed computer vision–based methods for PPE violation detection. However, limitations exist in implementing this in actual construction projects due to the costs associated with the acquisition and labeling of a large number of images, and the accuracy and efficiency of recording PPE violation data. Therefore, this study introduces semisupervised object detection (SS-OD) and data augmentation for training PPE detection models to reduce the use of labeled data without reducing the performance, and proposes a framework for recording the PPE violation data by extracting PPE violation videos from real-time surveillance footage by considering the role of the worker and their position, thereby enhancing practical construction management and serving as input for BBS-based safety risk assessments on construction sites. The results show that (1) the labeled data demand for training PPE detection models can be reduced through semisupervised learning, image augmentation, and transfer learning, without reducing the performance of PPE detection, thereby reducing the cost of application on actual construction sites; (2) SS-OD methods are better equipped to handle changes in construction scenarios by making full use of unlabeled data, and thus are suitable for construction scenarios; and (3) recording PPE violation data using the video clip method in a real construction project achieved an average precision of 91.76% and an average score of 92.79%. Using the PPE detection model trained with SS-OD effectively records PPE violation data for BBS-based safety risk assessment. This study significantly enhances the efficiency of real construction site PPE violation inspections and provides a valuable method for the automated and real-time collection of violation data in BBS-based management.
Computer vision–based personal protective equipment violation detection methods from past studies have been difficult to implement widely in real projects due to cost and effectiveness limitations. This study effectively improves upon these limitations using semisupervised learning and video clipping methods. Experimental results demonstrate that the object detection model trained with the semisupervised learning method not only reduces the need for labeled data in personal protective equipment detection model training but also exhibits improved generalization capabilities in complex and dynamic construction scenarios. Using the personal protective equipment violation video clipping method proposed in this study allows for efficient editing of violation video clips, facilitating rapid identification and documentation of construction workers’ personal protective equipment violation behaviors. This enhances the safety culture at construction sites and aids construction managers in developing safety management measures and educational plans. The real-time-collected personal protective equipment violation data can be utilized further as input for a behavior-based safety risk assessment model, enhancing the automation of unsafe behavior data collection.
Real-Time Detection of Personal Protective Equipment Violations for Construction Workers Using Semisupervised Learning and Video Clips
J. Constr. Eng. Manage.
Chen, Qihua (author) / Long, Danbing (author) / Wang, Siqi (author) / Chen, Qirong (author) / Yuan, Beifei (author)
2025-03-01
Article (Journal)
Electronic Resource
English
Real-Time Personal Protective Equipment Compliance Detection Based on Deep Learning Algorithm
| DOAJ | 2022