基于图像处理的隧道裂缝安全预警  被引量:9

Tunnel crack safety early warning based on image processing

在线阅读下载全文

作  者:陈立潮[1] 张媛媛 秦宇强 曹建芳 张锦[3,4] CHEN Li-chao;ZHANG Yuan-yuan;QIN Yu-qiang;CAO Jian-fang;ZHANG Jin(School of Computer Science and Technology,Taiyuan University of Science and Technology,Taiyuan 030024,China;Department of Computer,Xinzhou Teachers University,Xinzhou 034000,China;Department of Construction Machinery,Shanxi Traffic Vocational and Technical College,Taiyuan 030031,China;School of Mechanical Engineering,Taiyuan University of Science and Technology,Taiyuan 030024,China)

机构地区:[1]太原科技大学计算机科学与技术学院,山西太原030024 [2]忻州师范学院计算机系,山西忻州034000 [3]山西交通职业技术学院工程机械系,山西太原030031 [4]太原科技大学机械工程学院,山西太原030024

出  处:《计算机工程与设计》2020年第12期3479-3484,共6页Computer Engineering and Design

基  金:山西省交通厅科技项目计划基金项目(2019-1-9);山西省自然科学基金项目(201701D21059)。

摘  要:针对现今隧道裂缝检测准确度低与安全预警方面不足的问题,对隧道裂缝的检测算法、宽度计算方法及危害评级方法进行研究。提出一种利用PCA降维弱化噪声,泊松融合法加强裂缝对比度的裂缝分割法,并以此为基础设计一种将图像分块后测量投影点的欧氏距离的裂缝宽度计算方法。在CMOS相机获取的600张图像实验结果中,分割算法能精准检测隧道裂缝的区域,使宽度计算方法将裂缝宽度的估算值误差控制在5%左右,对裂缝进行有效的危害性评级与安全预警。To solve the problems of low accuracy of tunnel crack detection and insufficient safety early warning,the detection algorithm,width calculation method and hazard rating method of tunnel crack detection were studied.A crack segmentation method was proposed,in which PCA dimension reduction was used to weaken noise and Poisson fusion method was used to enhance crack contrast.On this basis,a crack width calculation method was designed to measure the Euclidean distance of projection points after image segmentation.Experimental results of 600 images obtained by CMOS camera show that,the segmentation algorithm can accurately detect the crack area,so that the width calculation method can control the estimated error of crack width to 5%,and effective hazard rating and safety early warning of cracks are carried out.

关 键 词:隧道裂缝 裂缝检测 裂缝宽度 危害等级 安全预警 

分 类 号:TP391[自动化与计算机技术—计算机应用技术]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象