Detecting safety helmet wearing on construction sites with bounding‐box regression and deep transfer learning

Abstract

Detecting safety helmet wearing in surveillance videos is an essential task for safety management, compliance with regulations, and reducing the death rate from construction industry accidents. However, it is much challenged by some factors like interocclusion, scale variances, perspective distortion, small object detection, and the carrier recognition of safety helmet. Traditional image‐based methods suffer from them. This article proposes a new methodology for detecting safety helmet wearing, which makes use of convolutional neural network‐based face detection and bounding‐box regression for safety helmet detection. On the one hand, the method can help to recognize the carrier of the safety helmet and detect a multiscale and small safety helmet. On the other hand, deep transfer learning based on DenseNet is introduced and applied using two different strategies, namely, object feature extractor and fine‐tuning for safety helmet recognition. To further improve the recognition accuracy, the network model with two peer DenseNet networks is trained by mutual distillation. Extensive analysis and experiments show that the novel methodology has considerable advantages in detecting safety helmet wearing compared to other state‐of‐the‐art models. The proposed model has achieved 96.2% recall, 96.2% precision, and 94.47% average detection accuracy. These results, precision‐recall (PR) curve, and receiver operating characteristic (ROC) curve demonstrate the feasibility of the new model.

Proactive proximity monitoring with instance segmentation and unmanned aerial vehicle‐acquired video‐frame prediction

“…H. Luo et al (2018) developed a CNN to identify the activities of workers, for example, walking, transporting, and stage bending, during the rebar installation process. Shen et al (2020) used a CNN for detecting safety helmets on construction sites with bounding-box regression and deep transfer learning. Arabi, Haghighat, and Sharma (2020) proposed vision-based constructionvehicle detectors with embedded devices based on the MobileNet.…”

“…Shen et al. (2020) used a CNN for detecting safety helmets on construction sites with bounding‐box regression and deep transfer learning. Arabi, Haghighat, and Sharma (2020) proposed vision‐based construction‐vehicle detectors with embedded devices based on the MobileNet.…”

Deep learning‐based object identification with instance segmentation and pseudo‐LiDAR point cloud for work zone safety

“…With the development of civil and infrastructure engineering (Shen et al., 2020), work zone safety problems are gaining more and more attention (X. Jiang & Adeli, 2004; Karim & Adeli, 2003; Raddaoui et al., 2020). Construction workers who died as a result of struck‐by‐vehicle account for approximately 57% of the total number of deaths in the construction work zone (Wang et al., 2017).…”

“…Currently, imaging sensors have been widely deployed on construction sites, and many algorithms have been proposed to detect various objects in videos and images and solve the problem of work zone safety. For example, detecting safety helmet wearing on construction sites reduces the death rate from construction industry accidents (Shen et al., 2020). Identifying various construction vehicles can be helpful to a safety monitoring system (Arabi et al., 2020).…”

Detection and Location of Safety Protective Wear in Power Substation Operation Using Wear-Enhanced YOLOv3 Algorithm

“…In [27], online hard example mining was combined with a multi-part detection method to identify whether a worker is wearing a safety helmet. Bounding-box regression and transfer learning was used to improve the convolutional neural network-based face detection for safety helmet detection in [28].…”