基于机器视觉技术的高层建筑施工现场危险区域识别方法  

Research on the Identification Method of Hazardous Areas in High Rise Building Construction Sites Based on Machine Vision Technology

作　　者：石雪洁 SHI Xuejie(Anhui Vocational and Technical College,Hefei 230012,China)

机构地区：[1]安徽职业技术学院,安徽合肥230012

出　　处：《佳木斯大学学报（自然科学版）》2023年第3期104-107,共4页Journal of Jiamusi University：Natural Science Edition

基　　金：安徽省重点自然科学项目(2022AH052067);安徽省重点自然科学项目(2022AH052076);2023年度安徽省重点自然科学项目。

摘　　要：针对传统方法对建筑施工现场危险区域识别不准确的问题,提出了基于机器视觉技术的高层建筑施工现场危险区域识别方法。首先利用视觉传感器采集施工现场图像,并将其存储于数据层;然后采用灵活性贝叶斯分类器建立危险区域识别方法,提取滤波后图像特征,计算特征所属危险区域的概率,以此识别含危险区域的施工现场图像;最后使用滑动区间方差检索灰度熵曲线突变范围,定位图像中危险区域坐标。实验结果显示:该方法应用在智慧工地管理系统后,可判断施工危险区域,且给出具体坐标位置。Aiming at the problem that the traditional method is not accurate in identifying the dangerous area of the construction site,a method of identifying the dangerous area of high-rise building construction site based on machine vision technology is proposed.Firstly,the image of construction site is collected by visual sensor and stored in the data layer.Then,the flexible Bayesian classifier is used to establish the hazard area recognition method,extract the filtered image features,and calculate the probability of the hazard area to which the features belong,so as to identify the construction site image containing the hazard area.Finally,sliding interval variance is used to retrieve the abrupt range of gray entropy curve and locate the coordinates of dangerous areas in the image.The experimental results show that this method can be used in the intelligent construction site management system to determine the construction danger area and give the specific coordinate position.

关 键 词：机器视觉技术 高层建筑 施工现场 危险区域识别 危险区域定位 

分 类 号：TU171[建筑科学—建筑理论]