检索规则说明:AND代表“并且”;OR代表“或者”;NOT代表“不包含”;(注意必须大写,运算符两边需空一格)
检 索 范 例 :范例一: (K=图书馆学 OR K=情报学) AND A=范并思 范例二:J=计算机应用与软件 AND (U=C++ OR U=Basic) NOT M=Visual
作 者:李昊朋 王景成[2] 黄姣茹 LI Haopeng;WANG Jingcheng;HUANG Jiaoru(School of Electronic Information Engineering,Xi'an University of Technology,Xi'an 710016;Department of Automation,Shanghai Jiao Tong University,Shanghai 200240)
机构地区:[1]西安工业大学电子信息工程学院,西安710016 [2]上海交通大学自动化系,上海200240
出 处:《计算机与数字工程》2022年第10期2176-2181,共6页Computer & Digital Engineering
基 金:国家自然科学基金重点项目“全断面大型掘进装备集成控制与优化运行理论及应用”(编号:61633019);陕西省重点研发计划项目“不确定环境下数据驱动的全流程建模与工艺参数优化”(编号:2021GY-067)资助。
摘 要:工厂内异常人员行为是造成一些人力资源浪费和安全事故发生的主要原因。论文针对此问题初步制定了5种决策融合策略,使用机器视觉的方法检测工厂人员的异常行为。首先使用YOLOv5算法对多人图像进行切割;之后利用Blaze-Pose算法计算分割图像中的单人关键点,以此作为特征使用角度识别和ST-SVM分类器识别分割后图像中人员的行为,分别得到两种图像行为信息;最后通过制定的决策融合策略,对两种图像行为进行决策融合,得到最终的人员行为信息。论文使用1000张图片作为数据集,进行试验;实验结果表明,文中方法可以准确识别常见的工厂人员异常行为,精确率大于0.80,召回率可达到89%以上,F1指数大于0.91,预期可以为工厂内人员异常行为检测与预警提供技术支持。The abnormal behave of factory personnel is the main reason for the waste of human resources and safety accidents.For this problem,this paper preliminarily formulates five decision fusion strategies,and uses the method of machine vision to detect the abnormal behavior of factory personnel. Firstly,the YOLOv5 algorithm is used to cut the multi person image. Then,the blaze pose algorithm is used to calculate the single key points in the segmented image,and then the angle recognition and ST-SVM classifier are used to recognize the behavior of the segmented single image,and two kinds of image behavior information are obtained respectively. Finally,through the decision fusion strategy,the two image behaviors are fused to obtain the final personnel behavior information. In this paper,1000 images are used as the data set for the experiment. The experimental results show that the method in this paper can accurately identify the common abnormal behavior of factory personnel,the accuracy rate is greater than 0.80,the recall rate can reach more than 89%,and the F1 index is greater than 0.91. It is expected to provide technical support for the detection and early warning of abnormal behavior of factory personnel.
关 键 词:机器视觉 目标检测 yolov5 Blaze Pose ST—SVM 行为识别
分 类 号:TP389.1[自动化与计算机技术—计算机系统结构]
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在载入数据...
正在链接到云南高校图书馆文献保障联盟下载...
云南高校图书馆联盟文献共享服务平台 版权所有©
您的IP:216.73.216.249