实验室保护设备识别算法基于改进的YOLOv7识别算法.
Huijuan Luo1,2, Wenjing Liu2, Pinghu Xu1
1National Center for Materials Service Safety, University of Science and Technology Beijing, Beijing, 100083, China.
Heliyon
|September 10, 2024
概括
改进的YOLOv7算法提高了实验室视频中小型个人防护设备 (PPE) 目标的检测. 这种智能系统通过在复杂的环境中准确识别缺失的头盔,护目镜和口罩来提高安全合规性.
科学领域:
- 计算机视觉 计算机视觉
- 人工智能的人工智能
- 实验室的安全工作.
背景情况:
- 实验室人员经常忽视安全标准,未能佩戴必要的个人防护设备 (PPE).
- 手动检查个人防护设备是否符合要求是低效的,昂贵的,容易出现错误的.
- 在视频监控中检测小型个人防护设备对现有算法构成重大挑战.
研究的目的:
- 为实验室个人防护设备 (PPE) 开发智能和高效的检测技术.
- 在复杂的实验室环境中提高个人防护设备识别的准确性和稳定性.
- 解决目前在视频监控中检测小目标的方法的局限性.
主要方法:
- 开发了一种改进的YOLOv7算法,将全球注意力机制 (GAM) 纳入高效层聚合网络 (ELAN) 进行增强的特征提取 (ELAN-G).
- 标准化高斯瓦瑟斯坦距离 (NWD) 度量被引入以改善小目标的检测,取代了CIoU度量.
- 为了训练和评估算法,构建了一个新的多维实验室个人防护设备 (PPE) 数据集.
主要成果:
- 改进的YOLOv7模型实现了平均平均精度 (mAP) 的84.2%,比原始模型增加了2.3%.
- 与基线相比,该算法显示检测率有5%的改善,Micro-F1得分有2%的改善.
- 实验结果显示,与当前算法相比,准确度显著提高,特别是在复杂场景中针对小型PPE目标.
结论:
- 建议改进的YOLOv7算法有效地解决了在复杂的实验室环境中检测小型个人防护设备 (PPE) 的挑战.
- 这种智能检测系统为提高实验室安全管理和合规性提供了强大的解决方案.
- 开发的方法显著提高了在现实实验室环境中识别PPE的准确性和效率.
相关概念视频
01:20Personal Protective Equipment
1.5K
Personal protective equipment (PPE) is unique clothing or equipment worn by an employee to minimize or prevent exposure to infectious agents. PPE creates a barrier between the employee and the infectious materials. PPE must be readily available in the patient care area. PPE includes gloves, gowns and aprons, masks and respirators, goggles, face shields, shoes, and headcovers:
1.5K
01:22PPE Use in Healthcare Settings I: Donning
939
Donning PPE must be completed before contact with the patient. This process protects from infectious agents. The sequence and action included in each donning are critical, and the steps must be systematic to avoid exposure to pathogens. The institutional policy also needs to be followed while donning PPE. The pre-donning preparations are gathering equipment, inspecting the PPE equipment for tears, holes, or damage, removing jewelry, removing any garments below the elbows, and tying the hair...
939