基于主动激励红外成像的大坝渗漏AI识别方法研究  

Study on dam leakage AI detection method based on infrared thermal imaging

作　　者：王嘉浩 丁勇[1] 黄英豪[2] 王羿 吴玉龙[3] WANG Jiahao;DING Yong;HUANG Yinghao;WANG Yi;WU Yulong(University of Science,Nanjing University of Science and Technology,Nanjing 210094,China;Nanjing Hydraulic Research Institute,Nanjing 210024,China;Kunshan Construction Engineering Quality Inspection Center,Kunshan Jiangsu 215337,China)

机构地区：[1]南京理工大学理学院,南京210094 [2]南京水利科学研究院,南京210024 [3]昆山市建设工程质量检测中心,江苏昆山215337

出　　处：《激光杂志》2024年第3期81-86,共6页Laser Journal

基　　金：中央级公益性科研院所基本科研业务费专项资金(No.Y322008);国家重点研发计划资助项目(No.2022YFC3005502);国家自然科学基金资助项目(No.51979174);国家自然科学基金联合基金项目(No.U2040221)。

摘　　要：为了解决大坝渗漏识别的问题,本文提出了一种将主动激励红外成像与深度学习结合的大坝渗漏识别方法。通过计算机仿真制作渗漏红外图像,再结合模拟大坝渗漏试验采集得到的红外图像,生成渗漏红外图像数据集用于深度学习的训练。在YOLOv5原始模型的基础上,用AF-FPN替换原有的FPN,提高识别大坝红外图像渗漏区域的能力,并在识别速度和准确率之间做出有效的权衡。试验表明,模型的准确率为87.6%,召回率为96.5%,平均准确率(IoU=0.5)为88.3%,表明本文提出的方法可较好的识别大坝红外图像渗漏区域。In order to solve the problem of dam leakage detection,this paper presents a dam leakage detection method which combines active excitation infrared imaging with depth learning.The infrared image of leakage is produced by computer simulation,and then combined with the infrared image acquired by simulating dam leakage test,the infrared image data set is generated for the training of depth learning.On the basis of Yolov5 original model,using AF-FPN to replace FPN can improve the ability of identifying the leakage area of dam infrared image,and make an effective trade-off between identifying speed and accuracy.The test results show that the accuracy of the model is 87.6%,the recall rate is 96.5%,and the average accuracy(IoU=0.5)is 88.3%,which indicates that the method proposed in this paper can identify the leakage area of dam infrared image well.

关 键 词：大坝 红外图像 主动激励 深度学习 渗漏识别 

分 类 号：TN209[电子电信—物理电子学]