基于视觉增强的桥梁检测算法及技术  

Visual enhancement-based bridge detection algorithm and technique

作　　者：朱尧于 李佳欢 朱力[2] 刘玉静 何超 刘涛 Zhu Yaoyu;Li Jiahuan;Zhu Li;Liu Yujing;He Chao;Liu Tao(CCCC Highway Bridges National Engineering Research Centre Co.,Ltd.,Beijing 100088,China;School of Civil Engineering,Beijing Jiaotong University,Beijing 100044,China;Jiangsu Expressway Engineering Maintenance Technology Co.,Ltd.,Nanjing 210049,China;Jiangsu Sutong Bridge Co.,Ltd.,Nantong 226017,China)

机构地区：[1]中交公路长大桥建设国家工程研究中心有限公司,北京100088 [2]北京交通大学土木建筑工程学院,北京100044 [3]江苏高速公路工程养护技术有限公司,南京210049 [4]江苏苏通大桥有限责任公司,南通226017

出　　处：《东南大学学报（自然科学版）》2024年第4期902-910,共9页Journal of Southeast University：Natural Science Edition

基　　金：国家青年人才计划“青年拔尖”资助项目;中交集团青年创新资助项目(2021-ZJKJ-QNCX03);中交集团院士专项资助项目(YSZX-03-2021-01-B)。

摘　　要：为了识别桥梁结构微小振动,提出一种基于方向码匹配(OCM)法的靶标圆心检测算法.该算法首先采用OCM法实现像素级匹配,然后在最佳匹配区域采用靶标圆心检测算法进行精确跟踪.通过室内试验,将基于OCM的靶标圆心检测算法与基于相位的光流法和OCM亚像素实现算法进行对比,研究3种算法在不同振动幅度和实际工程应用的适用性,为实桥检测提供技术方案比选依据.结果表明:所提算法在各种振动幅度下均能精确地识别结构振动位移和频率,在实桥应用时应考虑靶标粘贴的可行性;基于相位的光流法在幅度指标小于2 mm或1.7像素的微小振动工况下能精确地识别结构的振动位移和频率;OCM亚像素实现算法的分辨精度取决于插值点的数量,对大幅度振动工况的识别具有与所提算法相似的效果.In order to identify small vibrations of the bridge structure,a target center detection algorithm based on orientation code matching(OCM)method is proposed.The algorithm first uses the OCM method to achieve pixel-level matching,and then uses the target center detection algorithm for accurate tracking in the best matching area.Through laboratory experiments,the OCM-based target center detection algorithm is compared with the phase-based optical flow method and the OCM sub-pixel implementation algorithm,and the applicability of the three algorithms in different vibration amplitudes and practical engineering applications is explored,which provides a technical scheme comparison basis for real bridge detection.The results show that the proposed algorithm can accurately identify the structural vibration displacement and frequency under various vibration amplitudes,and the feasibility of target sticking should be considered in the application of real bridge.The phase-based optical flow method can accurately identify the vibration displacement and frequency of the structure under small vibration cases with amplitude index less than 2 mm or 1.7 pixels.The resolution accuracy of the OCM sub-pixel implementation algorithm depends on the number of interpolation points,and has a similar effect to the proposed algorithm under large amplitude vibration cases.

关 键 词：桥梁检测 适用性研究 视觉增强 方向码匹配 基于相位的光流法 

分 类 号：U44[建筑科学—桥梁与隧道工程]