墙体表面的网状裂纹最大宽度测量  

Measurement of the maximum width of mesh cracks on wall surface

作　　者：丁宁[1] 陈晓竹[1] 

机构地区：[1]中国计量大学信息工程学院,浙江杭州310018

出　　处：《中国计量大学学报》2017年第2期185-189,233,共6页Journal of China University of Metrology

摘　　要：建筑物墙体裂纹是重要的安全隐患,检测混凝土墙体表面的裂纹及测量其最大宽度,已引起众多关注.现介绍基于图像处理的智能检测方法,即根据裂纹像素点分布特征,利用连通域面积大小来提取裂纹,并删除伪裂纹等杂质,再对含有分支或网状裂纹进行局部处理,根据裂纹特征像素点的位置关系获取聚类近似初始值,之后利用K-means聚类算法不断迭代计算裂纹特征像素点到其对应直线的最短距离,并以此将图像中的像素点归为不同方向的裂纹类.最后,利用分类好的裂纹像素点分别进行边缘检测与最大宽度测量并比较,来获取含有交叉裂纹的最大宽度值.本文获得的水平裂纹最大宽度的相对误差为2.968%,斜垂裂纹最大宽度的相对误差为5.188%.Detection of concrete wall cracks and measurement of the maximum width of the cracks have attracted much attention since wall cracks could cause severe safety hazard. In this paper, an intelligent crack detection method based on image processing was introduced. According to the distribution of crack pixels, the crack was extracted by using the size of the connected domain, and the impurities such as pseudo crack were removed. According to the location relation of the pixels of the crack, the initial value of the cluster was obtained, and then the K means clustering algorithm was used to iteratively calculate the shortest distance between the feature points of the crack and the corresponding line and to classify the pixels in the image as cracks of different directions. The maximum width of the cross crack was obtained by using the categorized crack pixels to perform edge detection and maximum width measurement respectively. The relative error of the maximum crack width was 2. 968%, and the relative error of the maximum width of the slanting crack was 5. 188%.

关 键 词：裂纹检测 图像处理 K-MEANS聚类算法 特征提取 

分 类 号：TP306.2[自动化与计算机技术—计算机系统结构]