水工建筑物混凝土裂缝非接触高精度监测设备研发OACSTPCD
Research and development of non-contact high-precision monitoring equipment for concrete cracks in port construction buildings
目前水工建筑物裂缝检测仍以人力识别为主,常规仪器检测为辅,来实现一定频次的检测,效率低下且存在盲区.以水工建筑物混凝土裂缝为研究对象,对裂缝图像识别系统平台与监测设备进行深入研发.首先设计了混凝土裂缝非接触高精度监测设备构成,对于图像处理模块,通过利用深度学习U-Net 技术实现裂缝图像精准分割,借助切分处理操作,实现端对端的图片处理;然后基于中垂线的图像腐蚀技术计算裂缝宽度,基于霍夫变换去除噪声实现裂缝长度的测量.依据相关规范,将混凝土构件分为板、梁、桩与桩帽、方块、胸墙、墩台等构件,完善裂缝监测评级依据.最后测试了水工建筑物裂缝高精度监测设备工程样机的可靠性和精度.结果表明:水工建筑物混凝土裂缝非接触高精度监测设备对裂缝识别的精度可达0.08 mm,能够实现裂缝的自动识别、类型判断、精准测量、评级预警等功能,达到了水工建筑物结构开裂的智慧化监测效果.
At present,crack detection in hydraulic structures primarily relies on manual identification,supplemented by conventional instrument detection,achieving a certain frequency of detection that is inefficient and has blind spots.Taking concrete cracks in hydraulic structures as the research object,a crack image recognition system platform and monitoring equipment were comprehensively developed.Firstly,a non-contact high-precision monitoring device for concrete cracks was designed.For the image processing module,deep learning U-Net technology was used to achieve precise segmentation of crack images.With the help of segmentation processing operations,end-to-end image processing was achieved.Subsequently,based on the image erosion technique of the perpendicular line,the crack width was calculated,and the crack length was measured by removing noise using the Hough transform.According to relevant regulations,concrete components were divided into four categories:slabs,beams,piles and pile caps,blocks,breast walls,and piers.Furthermore,crack monitoring and rating criteria were improved.Finally,the reliability and accuracy of the engineering prototype of high-precision monitoring equipment for cracks in hydraulic structures were tested.On the surface of the results,the non-contact high-precision monitoring equipment for concrete cracks in hydraulic structures achieves an accuracy of 0.08 mm in crack recognition.This systems enables automatic identification,type classification,accurate measurement,rating warnings and other functions of cracks,achieving the intelligent monitoring effect of structural cracking in hydraulic structures.
朱鹏瑞;薛润泽;刘孟孟;王亚民;尹纪龙;张干;王鑫
交通运输部天津水运工程科学研究所 水工构造物检测、诊断与加固技术交通行业重点实验室,天津 300456天津理工大学 海洋能源与智能建设研究院,天津 300456
交通运输
水工建筑物混凝土裂缝裂缝识别深度学习监测设备
port buildingsconcrete crackscrack recognitionmachine visionmonitoring equipment
《水道港口》 2024 (004)
518-525 / 8
国家重点研发计划项目(2022YFB2603000);广西科技重大专项项目(桂科 AA23062045);浙江省科技项目(2022C01004);中央级公益性科研院所科研创新基金项目(TKS20230104)
评论