基于计算机视觉技术的裂缝特征参数提取方法研究与裂缝检测系统设计  

Method for Extraction of Crack Characteristic Parameters by Computer Vision Technology and Design of Crack Detection System

作　　者：顾征宇[1] 潘钻峰[2] 秦建宇 杨毅超[1] 李文迪 GU Zhengyu;PAN Zuanfeng;QIN Jianyu;YANG Yichao;LI Wendi(Shanghai Tobacco Group Co.,Ltd.,Shanghai 200082,China;State Key Laboratory of Disaster Reduction in Civil Engineering,Tongji University,Shanghai 200092,China)

机构地区：[1]上海烟草集团有限责任公司,上海200082 [2]同济大学土木工程防灾减灾全国重点实验室,上海200092

出　　处：《结构工程师》2023年第6期1-8,共8页Structural Engineers

基　　金：国家自然科学基金(52078368,52038010)。

摘　　要：目前国内外学者针对混凝土结构所提出的裂缝图像处理算法可移植性较差,还未有较为通用的算法。对目前常用的图像预处理算法及特征参数提取算法在混凝土裂缝图像识别方面的适用性进行了研究,从相机的拍摄和校正、图像预处理、裂缝特征参数提取等方面对比了不同的算法处理效果,并自主开发了一套混凝土裂缝检测系统。At present,the crack image processing algorithms proposed for concrete structures are less portable,and there is no general algorithm yet.The applicability of currently used image processing algorithms and feature parameter extraction algorithms in concrete crack image recognition is studied by comparing different algorithms in terms of camera calibration,image pre-processing,crack feature parameter extraction,etc.A concrete crack detection system is developed.

关 键 词：混凝土结构 裂缝检测 计算机视觉 图像处理 

分 类 号：TU37[建筑科学—结构工程] TP391.41[自动化与计算机技术—计算机应用技术]