Research on application of helmet wearing detection improved by YOLOv4 algorithm.

Aiming at the problem that the model of YOLOv4 algorithm has too many parameters and the detection effect of small targets is poor, this paper proposes an improved helmet fitting detection model based on YOLOv4 algorithm. Firstly, this model improves the detection accuracy of small targets by adding multi-scale prediction and improving the structure of PANet network. Then, the improved depth-separable convolution was used to replace the standard 3 × 3 convolution, which greatly reduced the model parameters without reducing the detection ability of the model. Finally, the k_means clustering algorithm is used to optimize the prior box. The model was tested on the self-made helmet dataset helmet_dataset. Experimental results show that compared with the safety helmet detection model based on Faster RCNN algorithm, the improved YOLOv4 algorithm has faster detection speed, higher detection accuracy and smaller number of model parameters. Compared with the original YOLOv4 model, the mAP of the improved YOLOv4 algorithm is increased by 0.49%, reaching 93.05%. The number of model parameters was reduced by about 58%, to about 105 MB. The model reasoning speed is 35 FPS. The improved YOLOv4 algorithm can meet the requirements of helmet wearing detection in multiple scenarios.

Safety Helmet Detection Based on YOLOv5 Driven by Super-Resolution Reconstruction.

High-resolution image transmission is required in safety helmet detection problems in the construction industry, which makes it difficult for existing image detection methods to achieve high-speed detection. To overcome this problem, a novel super-resolution (SR) reconstruction module is designed to improve the resolution of images before the detection module. In the super-resolution reconstruction module, the multichannel attention mechanism module is used to improve the breadth of feature capture. Furthermore, a novel CSP (Cross Stage Partial) module of YOLO (You Only Look Once) v5 is presented to reduce information loss and gradient confusion. Experiments are performed to validate the proposed algorithm. The PSNR (peak signal-to-noise ratio) of the proposed module is 29.420, and the SSIM (structural similarity) reaches 0.855. These results show that the proposed model works well for safety helmet detection in construction industries.

Impact resistance test system for the helmet based on a polyvinylidene fluoride piezoelectric sensor array.

As the most commonly used personal protection equipment (PPE) in various production activities, the impact resistance of the helmet is of great importance. Referred to the conventional experimental method, this study constructs a helmet impact resistance test system with a polyvinylidene fluoride (PVDF) sensor array. Compared with the traditional test method, this study installed PVDF sensors on the contact surface between the headform and the helmet. The stress and its distribution on the headform are measured directly, which is helpful to evaluate the impact resistance of the helmet more accurately and comprehensively. Finally, the intra-group correlation coefficient (ICC) and the coefficient of variation (CV) of peak pressure of the repeated test results are calculated to evaluate the reliability of the test system, which shows high reliability. The test system is helpful for optimization of the helmet production design and further related research.

Workshop Safety Helmet Wearing Detection Model Based on SCM-YOLO.

In order to overcome the problems of object detection in complex scenes based on the YOLOv4-tiny algorithm, such as insufficient feature extraction, low accuracy, and low recall rate, an improved YOLOv4-tiny safety helmet-wearing detection algorithm SCM-YOLO is proposed. Firstly, the Spatial Pyramid Pooling (SPP) structure is added after the backbone network of the YOLOv4-tiny model to improve its adaptability of different scale features and increase its effective features extraction capability. Secondly, Convolutional Block Attention Module (CBAM), Mish activation function, K-Means++ clustering algorithm, label smoothing, and Mosaic data enhancement are introduced to improve the detection accuracy of small objects while ensuring the detection speed. After a large number of experiments, the proposed SCM-YOLO algorithm achieves a mAP of 93.19%, which is 4.76% higher than the YOLOv4-tiny algorithm. Its inference speed reaches 22.9FPS (GeForce GTX 1050Ti), which meets the needs of the real-time and accurate detection of safety helmets in complex scenes.

Effect of helmet design on impact performance of industrial safety helmets.

Background: Comparative studies of different helmet designs are essential to determine differences in helmet performance. The present study comparatively evaluated the impact performance of hardhat helmets, climbing-style safety helmets, and helmets with novel rotation-damping technologies to determine if advanced designs deliver improved protection. Methods: Six helmet designs from three categories of safety helmets were tested: two traditional hardhat helmets (HH Type I, HH Type II), two climbing-style helmets (CS Web, CS Foam), and two helmets with dedicated rotation-damping technologies (MIPS, CEL). Helmets were first evaluated in impacts of 31 J energy representing a falling object according to standard Z89.1-2014. Subsequently, helmets were evaluated in impacts representing a fall by dropping a helmeted head-neck surrogate at 275 J impact energy. The resulting head kinematics were used to calculate the probability of sustaining a head or brain injury. Results: Crown impacts representative of a falling object resulted in linear acceleration of less than 50 g in all six helmet models. Compared to crown impacts, front, side and rear impacts caused a several-fold increase in head acceleration in all helmets except HH Type II and CEL helmets. For impacts representative of falls, all helmets except the CEL helmet exhibited significantly increased head accelerations and an increased brain injury probability compared to the traditional HH Type I hardhat. Neck compression was 35%-90% higher in the two climbing-style helmets and 80% higher in MIPS helmets compared to the traditional HH type I hardhat. Discussion: Contemporary helmets do not necessarily deliver improved protection from impacts and falls compared to traditional hardhat helmets.

SHEL5K: An Extended Dataset and Benchmarking for Safety Helmet Detection.

Wearing a safety helmet is important in construction and manufacturing industrial activities to avoid unpleasant situations. This safety compliance can be ensured by developing an automatic helmet detection system using various computer vision and deep learning approaches. Developing a deep-learning-based helmet detection model usually requires an enormous amount of training data. However, there are very few public safety helmet datasets available in the literature, in which most of them are not entirely labeled, and the labeled one contains fewer classes. This paper presents the Safety HELmet dataset with 5K images (SHEL5K) dataset, an enhanced version of the SHD dataset. The proposed dataset consists of six completely labeled classes (helmet, head, head with helmet, person with helmet, person without helmet, and face). The proposed dataset was tested on multiple state-of-the-art object detection models, i.e., YOLOv3 (YOLOv3, YOLOv3-tiny, and YOLOv3-SPP), YOLOv4 (YOLOv4 and YOLOv4pacsp-x-mish), YOLOv5-P5 (YOLOv5s, YOLOv5m, and YOLOv5x), the Faster Region-based Convolutional Neural Network (Faster-RCNN) with the Inception V2 architecture, and YOLOR. The experimental results from the various models on the proposed dataset were compared and showed improvement in the mean Average Precision (mAP). The SHEL5K dataset had an advantage over other safety helmet datasets as it contains fewer images with better labels and more classes, making helmet detection more accurate.

Effects of ventilation openings in industrial safety helmets on evaporative heat dissipation.

OBJECTIVES: This study aimed to investigate the effects of ventilation openings in commercial industrial safety helmets (ISHs) on evaporative heat dissipation. METHODS: Seven models of commercial ISH were examined quantitatively by a sweating thermal head manikin (SHM) with six separate zones. To simulate summer outdoor conditions, the measurements were done in a climate chamber, with the room temperature and relative humidity set at 34.0°C and 50%, respectively. The shell temperature of SHM was set at 34.0°C. Wind was blown from the front or left side at 1.0, 2.0, and 3.0 m/s. The necessary heat flux to keep the manikin skin temperature at 34.0°C was counted as evaporative heat dissipation in each zone. RESULTS: Openings at the front and back, and openings between the body and brim of the helmet played a significant role in increasing the heat flux in Forehead zone, but in all zones as a total, the effects were not significant. Heat flux for ISH with openings on both the right and left sides was not significantly different from that without openings. CONCLUSIONS: Our study utilizing SHM showed that ventilation openings on both the right and left sides or front and back sides of commercial ISHs were not significantly effective in increasing total evaporative heat dissipation under an equivalent temperature of ambience and manikin shell. Further improvements on ISH are needed to increase evaporative heat dissipation.

Knowledge, Attitude And Practice On The Usage Of Safety Helmet Among Oil Palm Harvesters

Safety helmet become vital personal protective equipment especially in the plantation in preventing the head from injury. This study evaluate the knowledge, attitude and practice on safety helmet usage among harvesters, the association between knowledge, attitude and practice of safety helmet usage with head injury; and the significant differences of the safety helmet practices before and after the intervention. A cross-sectional study was conducted among 109 harvesters in two oil palm plantation located in Selangor, Malaysia. A set of questionnaire was used to collect the socio demographic background data, knowledge, attitude and practice on the usage of safety helmet. An intervention program through tool box talk on proper usage of safety helmet also was given followed by an observation to look for the differences before and after the tool box promotion on the use of safety helmet. Result from the descriptive analysis showed high score for knowledge, fair score for the attitude and practice among harvesters. There is no association between knowledge (X2=2.733; p>0.05), attitude (X2=2.546; p>0.05) and practice (X2=2.473; p>0.05) with the head injury. The result also gave no significant differences (p>0.05) of the practices before and after the intervention. However, the trends showed decrease in number of practices after the intervention. This study reveals that the knowledge, attitude and practice are not a prominent indicator for head injury among harvesters.