Scholarlink

<p>The manufacturing and construction industries are the lifeblood of any prosperous economy. Employers must ensure that their employees work in a safe and healthy environment. The most effective way to protect employees is to eliminate risks at their source. This key point highlighted the need to improve industrial workplace safety. As a result, this proposed work aims to create a real-time scalable Computer Vision (CV)-based Deep Learning (DL) framework for detecting workers’ Personal Protective Equipment (PPE) compliance. Model for image classification and recognition for monitoring industrial PPE To determine which employees are wearing safety gear, it analyzes real-time video from high-resolution cameras. A helmet, gloves, and shoes are the items the solution looks for in a worker. Despite recent significant advancements in image processing technology, there are still some drawbacks regarding detection accuracy, calibration time, and installation cost. This research provides a best-fit model for worker sites based on six DL image detection models. Updates about the workers’ PPE kit will be available to safety officials via Industrial Internet of Things (IIoT) applications. By implementing the high-speed, high-accuracy YOLOv7 architecture, the framework detects PPE items such as helmets, gloves, and shoes from real-time video feeds, and the proposed work achieved 89% prediction accuracy.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Deep Learning and Industrial Internet of Things (IIoT) Approach for Detecting Worker’s Protective Equipment

  • 期刊论文

  • 发表日期：

    2025年7月23日

查看全文
  • Beulah Jackson,
  • A. Siva Kumar,
  • P. V. Gopirajan,
  • S. Asha,
  • David Neels Ponkumar

摘要

The manufacturing and construction industries are the lifeblood of any prosperous economy. Employers must ensure that their employees work in a safe and healthy environment. The most effective way to protect employees is to eliminate risks at their source. This key point highlighted the need to improve industrial workplace safety. As a result, this proposed work aims to create a real-time scalable Computer Vision (CV)-based Deep Learning (DL) framework for detecting workers’ Personal Protective Equipment (PPE) compliance. Model for image classification and recognition for monitoring industrial PPE To determine which employees are wearing safety gear, it analyzes real-time video from high-resolution cameras. A helmet, gloves, and shoes are the items the solution looks for in a worker. Despite recent significant advancements in image processing technology, there are still some drawbacks regarding detection accuracy, calibration time, and installation cost. This research provides a best-fit model for worker sites based on six DL image detection models. Updates about the workers’ PPE kit will be available to safety officials via Industrial Internet of Things (IIoT) applications. By implementing the high-speed, high-accuracy YOLOv7 architecture, the framework detects PPE items such as helmets, gloves, and shoes from real-time video feeds, and the proposed work achieved 89% prediction accuracy.