基于分水岭算法的矩形钢管混凝土构件刚度裂缝检测  

Stiffness Crack Detection of Concrete-filled Rectangular Steel Tubular Members Based on Watershed Algorithm

作　　者：祖庆芝 ZU Qingzhi(School of Architectural Engineering,Zhangzhou Institute of Technology,Zhangzhou 363000,China)

机构地区：[1]漳州职业技术学院建筑工程学院,福建漳州363000

出　　处：《太原学院学报（自然科学版）》2022年第1期1-6,共6页Journal of TaiYuan University:Natural Science Edition

基　　金：漳州职业技术学院2021年校级课题(ZZY2021B095)。

摘　　要：现有矩形钢管混凝土构件刚度裂缝检测方法存在刚度裂缝检测精准度较差的问题,无法满足现今建筑安全需求,故提出基于分水岭算法的矩形钢管混凝土构件刚度裂缝检测方法。获取刚度裂缝数字图像,并对其进行适当的预处理,以此为基础,基于分水岭算法对梯度图像进行分割,以图像分割结果为依据,标记连通区域,剔除干扰区域,提取连通区域特征,依据提取特征识别刚度裂缝目标,连接裂缝片段,制定刚度裂缝检测程序,将获取的矩形钢管混凝土构件数字图像输入制定程序中,即可实现矩形钢管混凝土构件刚度裂缝的检测。仿真实验可知:该方法刚度裂缝检测长度误差与宽度误差较小,具备更好的刚度裂缝检测精准度,能够发挥更好的应用效能。The existing stiffness crack detection methods for concrete-filled rectangular steel tubular members have the problem of poor stiffness crack detection accuracy,which can not meet the current construction safety requirements.Therefore,a stiffness crack detection method for concrete-filled rectangular steel tubular members based on watershed algorithm is proposed.The digital image of stiffness crack is obtained and properly preprocessed.On this basis,the gradient image is segmented by the watershed algorithm.Based on the image segmentation results,the connected area is marked;the interference area is eliminated.The connected area features are extracted;the stiffness crack target is identified according to the extracted features;the crack segments are connected,and the stiffness crack detection program is formulated.By inputting the acquired digital image of concrete-filled rectangular steel tubular members into the formulated program,the stiffness crack detection of concrete-filled rectangular steel tubular members can be realized.The simulation results show that the proposed method causes smaller length error and width error;it has better stiffness crack detection accuracy and thus can have a better application efficiency.

关 键 词：分水岭算法 矩形钢管 混凝土构件 刚度 裂缝检测 

分 类 号：TU317[建筑科学—结构工程]