YOLO-NAS based Intelligent Workplace Safety Monitoring System for Real-Time Personal Protective Equipment Compliance Detection

AI-Driven Detection of Personal Protective Equipment Usage in Occupational Settings: Enhancing Workplace Safety

The primary objective of this study is to assess the effect of systematic dataset augmentation on the accuracy of real-time, vision-based Personal Protective Equipment (PPE) detection systems in occupational environments. The PPEDS-1000 dataset was employed, comprising 1,000 expertly annotated images across four PPE usage categories: worker (W), worker with helmet (WH), worker with vest (WV), and worker with both helmet and vest (WHV). An augmented dataset (PPEDS-2600) was derived via controlled geometric transformations (horizontal and vertical flips), additive Gaussian noise, and Gaussian blur. Each dataset is partitioned using an 80/10/10 train–validation–test split and utilized to train five YOLOv8 model variants (nano through extra-large). The evaluation metrics include precision, recall, F1-score, mean average precision at an IoU threshold of 0.5 (mAP50), and mean average precision averaged over IoU thresholds from 0.5 to 0.95 (mAP50-95). The experimental results demonstrate that augmentation elevates mAP50 from 77.7% on PPEDS-1000 to 94.8% on PPEDS-2600, thereby substantiating the hypothesis that targeted augmentation markedly enhances detection performance. The findings indicate that the present work establishes a rigorous benchmark for real-time PPE compliance monitoring.

AN OPTIMIZED YOLOV10 FRAMEWORK FOR DUAL DETECTION OF SAFETY HELMETS AND LICENSE PLATES

This research uses the most recent YOLOv10 object detection architecture to demonstrate a sophisticated computer vision system for real-time safety helmet detection and license plate recognition. The main goal is to improve vehicle monitoring and workplace safety by automatically recognising people who are not wearing safety helmets in industrial zones and recording license plates for regulatory and surveillance purposes. Efficient multi-object identification in difficult situations is made possible by YOLOv10, which is renowned for its exceptional speed and accuracy. To ensure reliable performance, the system is trained using annotated datasets that include a variety of helmet types and car plates under various circumstances. Construction workers' risk of suffering head injuries in highaltitude falls can be significantly decreased by donning safety helmets. This study suggests an enhanced safety helmet detection method based on YOLOv10 to solve the low detection accuracy of current algorithms for small objects and complicated settings in different situations.

Identification of fastening compliance for intelligent work-at-height safety belts based on multimodal time-series analysis

Working-at-height safety is a critical concern in domains such as industrial production and power maintenance. Traditional safety belts only offer passive protection and lack real-time monitoring of proper wearing and scenario suitability, which may cause accidents due to non-standard fastening or environmental misjudgment. To address this issue, this paper develops an intelligent safety belt model with self-identification and self-sensing capabilities. The model combines multisensor fusion techniques with temporal analysis models to achieve precise recognition of working-at-height scenarios and real-time assessment of safety belt fastening compliance. A collaborative fusion method for multi-source sensor data is established. Recurrent neural networks are used to analyze the time-series data collected from the various sensors, capturing the temporal dependencies of action sequences. Convolutional neural networks are employed to extract spatial features from proximity-sensor outputs, enabling accurate assessment of the spatial correctness of safety belt attachment points. An attention-based fusion prediction is applied to combine temporal and spatial representations, enabling accurate assessment of safety belt fastening compliance during working-at-height operations. Experimental results show that the proposed model significantly improves safety supervision for work at height. It also provides proactive protection for personnel performing elevated tasks.

Computer Vision-Based Smart Camera for Safety Helmet Detection in Work Areas

The application of computer vision technology in automation systems plays a crucial role in improving the efficiency of occupational safety monitoring in industrial environments. This study developed a YOLOv8-based visual detection application in ONNX format to identify safety helmet violations in real-time. The system was developed using Python with a Tkinter-based user interface and integrated with a Flask web dashboard that displays violation log data. The application can accept video input from various sources, including webcams, USB cameras, and IP cameras, to classify the type of helmet being used. Only orange and white safety helmets are considered valid. Detecting a new helmet, a motorcycle helmet, or a helmet with an inappropriate colour will trigger an alarm and store the image as evidence of the violation. The YOLOv8 model was trained on a six-class dataset and demonstrated good performance, with a precision of 0.921, a recall of 0.859, an mAP50 value of 0.919, and an mAP50-95 value of 0.619. System evaluation demonstrated the application's stability and accuracy in computer vision-based automated surveillance.

Smart Traffic Enforcement: A Deep Learning-based Framework for Helmet Violation Detection and Fine Collection

Research on safety helmet wearing detection algorithm for workers at water conservancy construction sites based on the HDS-DETR network

The construction safety of workers in hydraulic construction sites that are crowded and difficult to manage is very serious. When personnel movement is frequent, the status of workers wearing safety helmets is difficult to monitor in real time. The focus of this study is the design of HDS-DETR model which is aimed to improve the safety recognition in hydraulic construction projects. Improvements were achieved by integrating the C2f-HDRAB Module to the RT-DETR model to strengthen the model's ability to detect features, the D-Attention mechanism to improve the model's ability to recognize important features, and SlimNeck architecture was implemented to improve the model's ability to efficiently fuse features. The results of the experiments reflect that the accuracy achieved was 94.1% and 89.6% of the improved model offered by the dedicated dataset in recall, and 94.9% of the mean Average Precision at IoU threshold 0.5, which is a 3.7% increase in the original model. The ablation tests demonstrate the effectiveness of the correction of modules and the proposed design is aimed at the complex nature of hydraulic construction, and provides real-time hard hat wearing monitoring. Safety management of the hydraulic engineering construction project provides support and improves the safety condition recognition in smart water conservancy construction projects.