基于机器视觉的桥梁缺陷自动识别技术研究  

RESEARCH ON AUTOMATIC RECOGNITION TECHNOLOGY OF BRIDGE DEFECTS BASED ON MACHINE VISION

作　　者：赵琳 Zhao Lin

机构地区：[1]上海市建筑科学研究院有限公司,上海201108

出　　处：《建筑技术开发》2024年第8期78-81,共4页Building Technology Development

基　　金：2022年上海市科委优秀技术带头人计划项目(22XD1433300)。

摘　　要：本文围绕机器视觉技术对桥梁缺陷进行自动识别,旨在提升桥梁的安全性和可靠性。探讨数据采集与图像预处理、特征提取与选择、机器学习模型训练等关键环节,重点分析了桥梁缺陷自动识别的流程及区域划分,并提出了技术性能要求与可靠性考虑。通过采集和处理大量桥梁图像数据,采用高级图像处理技术和机器学习算法,实现了对裂缝、腐蚀和变形等缺陷的准确识别和定位,分析实际应用中遇到的挑战和解决方案,为桥梁检测和维护提供了科学的技术支持。The structural depth of Fengyiqiao South Station of Beijing Metro Fangshan Line is mostly below the groundwater level,so it must be dehydrated to achieve the purpose of waterless construction during construction.In order to effectively protect and utilize groundwater resources,the water resources extracted by precipitation are pumped back into the same underground water layer.This paper explores the influence of precipitation recharge on the settlement of pebble formation.Through field recharge water monitoring,surface settlement monitoring,theoretical calculation and numerical analysis calculation,it is proved that the theoretical calculation and numerical analysis of settlement can provide more accurate data support for the analysis of formation settlement change of precipitation recharge.According to the analysis results,it can be concluded that the surface settlement caused by precipitation recharge in the pebble formation is small and has almost no effect on the surrounding environment.

关 键 词：机器视觉 桥梁缺陷 自动识别技术 

分 类 号：TG441.7[金属学及工艺—焊接]