Real-Time Safety Helmet Detection Using Yolov5 at Construction Sites

The construction industry has always remained the economic and social backbone of any country in the world where occupational health and safety(OHS)is of prime importance.Like in other developing countries,this industry pays very little,rather negligible attention to OHS practices in Pakistan,resulting in the occurrence of a wide variety of accidents,mishaps,and near-misses every year.One of the major causes of such mishaps is the non-wearing of safety helmets(hard hats)at construction sites where falling objects from a height are unavoid-able.In most cases,this leads to serious brain injuries in people present at the site in general and the workers in particular.It is one of the leading causes of human fatalities at construction sites.In the United States,the Occupational Safety and Health Administration(OSHA)requires construction companies through safety laws to ensure the use of well-defined personal protective equipment(PPE).It has long been a problem to ensure the use of PPE because round-the-clock human monitoring is not possible.However,such monitoring through technological aids or automated tools is very much possible.The present study describes a systema-tic strategy based on deep learning(DL)models built on the You-Only-Look-Once(YOLOV5)architecture that could be used for monitoring workers’hard hats in real-time.It can indicate whether a worker is wearing a hat or not.The proposed system usesfive different models of the YOLOV5,namely YOLOV5n,YOLOv5s,YOLOv5 m,YOLOv5l,and YOLOv5x for object detection with the support of PyTorch,involving 7063 images.The results of the study show that among the DL models,the YOLOV5x has a high performance of 95.8%in terms of the mAP,while the YOLOV5n has the fastest detection speed of 70.4 frames per second(FPS).The proposed model can be successfully used in practice to recognize the hard hat worn by a worker.

HWD-YOLO:A New Vision-Based Helmet Wearing Detection Method

It is crucial to ensure workers wear safety helmets when working at a workplace with a high risk of safety accidents,such as construction sites and mine tunnels.Although existing methods can achieve helmet detection in images,their accuracy and speed still need improvements since complex,cluttered,and large-scale scenes of real workplaces cause server occlusion,illumination change,scale variation,and perspective distortion.So,a new safety helmet-wearing detection method based on deep learning is proposed.Firstly,a new multi-scale contextual aggregation module is proposed to aggregate multi-scale feature information globally and highlight the details of concerned objects in the backbone part of the deep neural network.Secondly,a new detection block combining the dilate convolution and attention mechanism is proposed and introduced into the prediction part.This block can effectively extract deep featureswhile retaining information on fine-grained details,such as edges and small objects.Moreover,some newly emerged modules are incorporated into the proposed network to improve safety helmetwearing detection performance further.Extensive experiments on open dataset validate the proposed method.It reaches better performance on helmet-wearing detection and even outperforms the state-of-the-art method.To be more specific,the mAP increases by 3.4%,and the speed increases from17 to 33 fps in comparison with the baseline,You Only Look Once(YOLO)version 5X,and themean average precision increases by 1.0%and the speed increases by 7 fps in comparison with the YOLO version 7.The generalization ability and portability experiment results show that the proposed improvements could serve as a springboard for deep neural network design to improve object detection performance in complex scenarios.

Real-time Safety Helmet-wearing Detection Based on Improved YOLOv5

Safety helmet-wearing detection is an essential part of the intelligentmonitoring system. To improve the speed and accuracy of detection, especiallysmall targets and occluded objects, it presents a novel and efficient detectormodel. The underlying core algorithm of this model adopts the YOLOv5 (YouOnly Look Once version 5) network with the best comprehensive detection performance. It is improved by adding an attention mechanism, a CIoU (CompleteIntersection Over Union) Loss function, and the Mish activation function. First,it applies the attention mechanism in the feature extraction. The network can learnthe weight of each channel independently and enhance the information dissemination between features. Second, it adopts CIoU loss function to achieve accuratebounding box regression. Third, it utilizes Mish activation function to improvedetection accuracy and generalization ability. It builds a safety helmet-wearingdetection data set containing more than 10,000 images collected from the Internetfor preprocessing. On the self-made helmet wearing test data set, the averageaccuracy of the helmet detection of the proposed algorithm is 96.7%, which is1.9% higher than that of the YOLOv5 algorithm. It meets the accuracy requirements of the helmet-wearing detection under construction scenarios.

Algorithm of Helmet Wearing Detection Based on AT-YOLO Deep Mode

The existing safety helmet detection methods are mainly based on one-stage object detection algorithms with high detection speed to reach the real-time detection requirements,but they can’t accurately detect small objects and objects with obstructions.Therefore,we propose a helmet detection algorithm based on the attention mechanism(AT-YOLO).First of all,a channel attention module is added to the YOLOv3 backbone network,which can adaptively calibrate the channel features of the direction to improve the feature utilization,and a spatial attention module is added to the neck of the YOLOv3 network to capture the correlation between any positions in the feature map so that to increase the receptive field of the network.Secondly,we use DIoU(Distance Intersection over Union)bounding box regression loss function,it not only improving the measurement of bounding box regression loss but also increases the normalized distance loss between the prediction boxes and the target boxes,which makes the network more accurate in detecting small objects and faster in convergence.Finally,we explore the training strategy of the network model,which improves network performance without increasing the inference cost.Experiments show that the mAP of the proposed method reaches 96.5%,and the detection speed can reach 27 fps.Compared with other existing methods,it has better performance in detection accuracy and speed.

GL-YOLOv5: An Improved Lightweight Non-Dimensional Attention Algorithm Based on YOLOv5

Craniocerebral injuries represent the primary cause of fatalities among riders involved in two-wheeler accidents;nevertheless,the prevalence of helmet usage among these riders remains alarmingly low.Consequently,the accurate identification of riders who are wearing safety helmets is of paramount importance.Current detection algorithms exhibit several limitations,including inadequate accuracy,substantial model size,and suboptimal performance in complex environments with small targets.To address these challenges,we propose a novel lightweight detection algorithm,termed GL-YOLOv5,which is an enhancement of the You Only Look Once version 5(YOLOv5)framework.This model incorporates a Global DualPooling NoReduction Blend Attention(GDPB)module,which optimizes the MobileNetV3 architecture by reducing the number of channels by half and implementing a parallelized channel and spatial attention mechanism without dimensionality reduction.Additionally,it replaces the conventional convolutional layer with a channel shuffle approach to overcome the constraints associated with the Squeeze-and-Excitation(SE)attention module,thereby significantly improving both the efficiency and accuracy of feature extraction and decreasing computational complexity.Furthermore,we have optimized the Variable Normalization and Attention Channel Spatial Partitioning(VNACSP)within the C3 module of YOLOv5,which enhances sensitivity to small targets through the application of a lightweight channel attention mechanism,substituting it for the standard convolution in the necking network.The Parameter-Free Spatial Adaptive Feature Fusion(PSAFF)module is designed to adaptively modify the weights of each spatial position through spatial pooling and activation functions,thereby effectively enhancing the model’s ability to perceive contextual information over distances.Ultimately,GL-YOLOv5 performs remarkably in the custom dataset,achieving a model parameter count of 922,895 M,a computational load of 2.9 GFLOPS,and a mean average precision(mAP)of 92.1%.These advancements significantly improve the model’s detection capabilities and underscore its potential for practical applications.