视觉驱动的涉水混凝土结构水下多类别缺陷辨识和区域弱监督定位方法  

Vision-Driven Underwater Multi-Category Defect Identification and Regional Weak Supervision Positioning Method for Wading Concrete Structures

作　　者：李扬涛 赵海涛[3] 魏洋[1] 包腾飞[2] 朱延涛[2] 王秋东 赵梦凡 LI Yangtao;ZHAO Haitao;WEI Yang;BAO Tengfei;ZHU Yantao;WANG Qiudong;ZHAO Mengfan(School of Civil Engineering,Nanjing Forestry University,Nanjing 210037,China;National Key Laboratory of Water Disaster Prevention,Hohai University,Nanjing 210098,China;College of Civil Engineering and Transportation,Hohai University,Nanjing 210098,China)

机构地区：[1]南京林业大学土木工程学院,江苏南京210037 [2]河海大学水灾害防御全国重点实验室,江苏南京210098 [3]河海大学土木与交通学院,江苏南京210098

出　　处：《应用基础与工程科学学报》2025年第1期60-75,共16页Journal of Basic Science and Engineering

基　　金：国家杰出青年科学基金项目(52325803);国家自然科学基金项目(52378244,52309152,U23B20150);江苏省自然科学基金项目(BK20231293,BK20220978);江苏省农业科技自主创新项目(CX(22)3183);国家基金委区域创新联合基金重点项目(U22A20229);淮河入海水道二期工程枢纽建设质量及安全保障提升关键技术研究(RHSD2/FW-2024-03);广西科技计划项目(桂科AA23062034);国网江苏省电力有限公司省管产业科技项目(JC2024119)。

摘　　要：大坝、桥梁等涉水建筑物水下混凝土结构缺陷的隐蔽程度高,常规人工检测和工程物探手段难以准确识别.基于此,在深入分析水下混凝土结构病害成因及类型基础上,融合现场搜集、开源信息获取和物理模型试验等多种手段,构建了工程结构缺陷图像数据库SD-ImageNet.为降低深度学习缺陷辨识模型的建模成本和数据依赖,结合域间和域内迁移学习策略,研究并提出了一种两阶段融合迁移学习策略,实现了混凝土结构通用图像特征提取.结合深度残差网络ResNet50、弱监督动态可视化理论与两阶段混合迁移学习策略,提出一种视觉驱动的涉水混凝土结构水下缺陷辨识方法,实现了多类别缺陷辨识和损伤区域弱监督定位.以某碾压混凝土重力坝为工程实例,从定性和定量两个维度评估了提出方法的适应性和有效性.Affected by obstacles such as reservoir water coverage and plankton,the underwater concrete structural defects of water-related buildings are highly concealed,making it difficult to accurately identify them with conventional manual inspection and engineering geophysical prospecting methods.As a carrier of high-dimensional data information,video images contain a large amount of important information closely related to the evolution of structural behavior such as the health status of the structure,mechanical evolution rules,and response to load and environmental effects.Based on an in-depth analysis of the causes and types of underwater concrete structural diseases,this paper proposes and constructs an engineering structural defect image database SD-ImageNet by integrating on-site collection,open source information acquisition,physical model testing,and other means.Combining inter-domain and intra-domain transfer learning strategies,a two-stage fusion transfer learning strategy was studied and proposed to achieve universal image feature extraction of concrete structures.On this basis,combined with the deep residual network ResNet50,weakly supervised dynamic visualization theory,and two-stage hybrid transfer learning strategy,a vision-driven underwater defect identification method for wading concrete structures is proposed to achieve multi-category defect identification and weak supervision of damaged areas position.Taking a roller-compacted concrete gravity dam as an engineering example,the adaptability and effectiveness of the proposed method are evaluated from both qualitative and quantitative dimensions.

关 键 词：深水检测 缺陷检测 病害识别 图像分类 人工智能 深度学习 

分 类 号：TU528[建筑科学—建筑技术科学]