摘要
混凝土裂缝的出现和发展将严重影响结构的安全性与适用性,因此对混凝土裂缝的有效检测尤为重要。传统的裂缝检测算法存在抗干扰能力弱、裂缝提取不完整等问题。提出一种结合形态学操作与对比度增强的预处理算法,以及进行阈值分割后基于裂缝几何特征的滤波算法,可以有效地实现混凝土裂缝的自动检测。预处理算法可以去除低对比度以及光照不均匀的影响,并且可以平滑背景噪声以及增强裂缝特征;基于裂缝几何特征的滤波算法可以利用面积特征、线性特征、凸性特征实现对裂缝区域的提取。试验表明,该算法可以较为精确的检测并提取出混凝土裂缝特征。
The appearance and development of concrete cracks will seriously affect the safety and applicability of the structure,therefore,the effective detection of concrete cracks is particularly important.The traditional crack detection algorithm has the problems of weak anti-interference ability and incomplete crack extraction.This paper proposes a preprocessing algorithm that combines morphological operations with contrast enhancement,and a filtering algorithm based on geometrical features of cracks,which can effectively achieve automatic detection of concrete cracks.The preprocessing algorithm can remove the effects of low contrast and uneven illumination,and can smooth background noise and enhance crack characteristics.The filtering algorithm based on the geometrical features of fractures can use the area features,linear features,and convex features to extract the fracture areas.The experiment result shows that the algorithm can detect and extract concrete crack features more accurately.
作者
李国耀
王腾
LI Guoyao;WANG Teng(Sun Yueqi Honors College,China University of Mining and Technology,Xuzhou 221116,China;School of Mechanics and Civil Engineering,China University of Mining and Technology,Xuzhou 221116,China)
出处
《建筑结构》
CSCD
北大核心
2020年第S02期529-533,共5页
Building Structure
基金
大学生创新训练计划项目(201810290159X)
关键词
混凝土裂缝
形态学处理
阈值分割
裂缝几何特征
concrete crack
morphological processing
threshold segmentation
crack geometry