基于AIOT 技术的深基坑一体化作业安全视觉监测方法  

Safety Visual Monitoring Method for Integrated Operation of Deep Foundation Pit Based on AIOT Technology

作　　者：刘少博 杜水婷 徐鹏 张建锋 LIU Shaobo;DU Shuiting;XU Peng;ZHANG Jianfeng(State Grid Gansu Electric Power Company,Lanzhou,Gansu 730050,China)

机构地区：[1]国网甘肃省电力公司,兰州甘肃730050

出　　处：《计算技术与自动化》2024年第2期41-46,共6页Computing Technology and Automation

摘　　要：针对深基坑一体化作业安全风险较高的问题,研究了基于AIOT技术的深基坑一体化作业安全视觉监测方法。利用双目视觉相机采集深基坑一体化作业监测区域图像,利用物联网节点将所采集图像传送至远程监测中心。远程监测中心接收图像后,利用卡尔曼滤波算法优化深基坑一体化作业监测图像。利用完成训练的YOLOv3算法,识别是否存在未佩戴安全帽以及误入危险区域的作业人员,存在未佩戴安全帽以及误入危险区域的作业人员时,利用物联网节点定位方法定位危险作业人员位置,实现深基坑一体化作业安全监测。实验结果表明,该方法有效监测深基坑一体化作业中,未佩戴安全帽以及进入危险区域的作业人员,提升深基坑一体化作业的管控水平。In view of the high safety risk of deep foundation pit integrated operation,a visual safety monitoring method for deep foundation pit integrated operation based on aiot technology is studied.The binocular vision camera is used to collect the image of the integrated operation monitoring area of deep foundation pit,and the Internet of things node is used to transmit the collected image to the remote monitoring center.After the remote monitoring center receives the images,the Kalman filter algorithm is used to optimize the monitoring images of deep foundation pit integrated operation.Use the YOLOv3 algorithm that has completed the training to identify whether there are operators who do not wear safety helmets and enter the dangerous area by mistake.When there are operators who do not wear safety helmets and enter the dangerous area by mistake,use the Internet of things node positioning method to locate the position of dangerous operators,so as to realize the integrated operation safety monitoring of deep foundation pit.The experimental results show that this method can effectively monitor the operators who do not wear safety helmets and enter the dangerous area in the integrated operation of deep foundation pit,and improve the control level of the integrated operation of deep foundation pit.

关 键 词：AIOT技术 深基坑 一体化 作业安全 视觉监测方法 YOLOv3 

分 类 号：TU753[建筑科学—建筑技术科学]