[1]王娟,申祖晨,姚远,等.基于计算机视觉的斗栱转动变形检测方法[J].西安建筑科技大学学报(自然科学版),2024,56(05):669-678.[doi:10.15986.j.1006-7930.2024.05.005]
　WANG Juan,SHEN Zuchen,YAO Yuan,et al. Rotation deformation detection method for Dou-Gong based on computer vision [J].J. Xi’an Univ. of Arch. & Tech.(Natural Science Edition),2024,56(05):669-678.[doi:10.15986.j.1006-7930.2024.05.005]
点击复制

基于计算机视觉的斗栱转动变形检测方法()
分享到：

西安建筑科技大学学报（自然科学版）[ISSN:1006-7930/CN:61-1295/TU]

卷:
56
期数:
2024年05期
页码:
669-678
栏目:
出版日期:
2024-10-28

文章信息/Info

Title:
Rotation deformation detection method for Dou-Gong based on computer vision
文章编号:
1006-7930(2024)05-0669-10
作者:
王娟1; 2; 申祖晨1; 2; 姚远1; 2; 杨娜1; 2
（1北京交通大学 土木建筑工程学院，北京 100044; 2北京交通大学 结构风工程与城市风环境北京市重点实验室，北京 100044）
Author(s):
（1School of Civil Engineering, Beijing Jiaotong University, Beijing 100044, China; 2Beijing′s Key Laboratory of Structural Wind Engineering and Urban Wind Environment, Beijing Jiaotong University, Beijing 100044, China）
关键词:
斗栱节点; 转动变形; 姿态估计; YOLOv8Pose; 直线检测
Keywords:
分类号:
TU366
DOI:
10.15986.j.1006-7930.2024.05.005
文献标志码:
A
摘要:
古建筑木结构斗栱节点由斗、栱等构件交错层叠而成.受自身形制、环境和外力等影响，这种“层叠式”节点易出现转动变形.针对人工及布设传感器的传统检测方法在可实施性及检测效率方面的局限性，提出了一种基于姿态估计的斗栱转动变形计算机视觉检测方法.首先定义了斗栱的关键点，并基于关键点推导出了各层斗的转动以及栌斗和阑额相对转动的计算公式；其次通过采集斗栱实景图像、实验室模型图像以及缩尺模型图像，构建了斗栱节点多样性数据集；而后搭建YOLOv8Pose姿态估计模型，并开展了6种规模和Batch Size的23种工况对比实验；结果表明，最优性能模型目标检测的mAP50(B)达到094，关键点检测的mAP50(P)达到091.最后利用缩尺模型转动变形检测实验验证了所提方法的有效性.
Abstract:
DouGong joints of ancient wooden structure are composed of Dou and Gong and other components. Affected by its own shape, environment and external force, such “stackable” joints are prone to rotation deformation and other types of damage. In view of the limitations of traditional detection methods of manual and deployed sensors in the implementation and detection efficiency, a computer vision detection method of Dougong rotation deformation based on pose estimation was proposed. Firstly, the key points of the DouGong were defined, and based on the key points, the calculation formula of the rotation of each layer of the DouGong and the relative rotation of the LuDou and the LanE were deduced. Secondly, a diverse dataset of DouGong joints was constructed by collecting Dougong real scene image, laboratory model image and scaled model image. Then, the YOLOv8Pose pose estimation model was built, and the comparative experiments were conducted under 23 conditions with 6 varying sizes and Batch Sizes. The results showed that the target detection part mAP50(B) of the optimal performance model reached 094, and the key point detection part mAP50(P) reached 091. The effectiveness of the proposed method was verified by the rotation deformation detection experiment of the scale model.

参考文献/References:

［1］隋䶮,赵鸿铁,薛建阳,等.古建木构枓栱侧向刚度的试验研究［J］.西安建筑科技大学学报(自然科学版), 2009,41(5): 668671.
SUI Yan, ZHAO Hongtie, XUE Jianyang, et al. Experimental research on the lateral stiffness of ancient wooden dougong［J］. J. Xi′an Univ. of Arch. & Tech. (Natural Science Edition), 2009,41(5): 668671.
［2］吴晨伟,薛建阳,宋德军,等.歪闪平身科斗栱轴压性能试验研究及有限元分析［J/OL］.工程力学, 2023: 111.20230613. ［20240426］. http://kns.cnki.net/kcms/detail/11.2595.O3.20230612.1556.003.html.
WU Chenwei, XUE Jianyang, SONG Dejun, et al. Experimental research and finite element analysis on axial compression performance of inclined pingshenke DOUGONG brackets［J/OL］. Engineering Mechanics, 2023: 111.20230613. ［20240426］.http://kns.cnki.net/kcms/detail/11.2595.O3.20230612.1556.003.html.
［3］邵云,邱洪兴,乐志,等.宋、清式斗栱低周反复荷载试验研究［J］. 建筑结构, 2014, 44(9): 7982, 38.
SHAO Yun, QIU Hongxing, LE Zhi, et al. Experimental study of lowcycle loading test on Songstyle and Qingstyle dougong［J］. Building Structure,2014,44(9): 7982, 38.
［4］谢启芳,向伟,杜彬,等.残损古建筑木结构叉柱造式斗栱节点抗震性能退化规律研究［J］.土木工程学报,2014,47(12): 4955.
XIE Qifang, XIANG Wei, DU Bin, et al. Study on seismic behavior degradation of damaged “Dougong” joint built with fork column of Chinese ancient timber structure［J］. China Civil Engineering Journal, 2014, 47(12): 4955.
［5］陈九璋, 陈雪瑶, 戴璐. 残损斗栱节点受力性能试验研究［J］. 北京林业大学学报, 2020,42(2): 149158.
CHEN Jiuzhang, CHEN Xueyao, DAI Lu. Experimental research on mechanical performance of damaged bracket set joints ［J］. Journal of Beijing Forestry University, 2020, 42(2): 149158.
［6］薛建阳,宋德军,吴晨伟,等.不同倾斜角度下平身科斗栱抗震性能研究［J/OL］. 工程力学, 2023: 113. 20231127. ［20240426］.http://kns.cnki.net/kcms/detail/11.2595.O3.20231124.1506.010.html.
XUE Jianyang, SONG Dejun, WU Chenwei, et al. Study on seismic behavior of DOUGONG bracket sets between columns under different inclinations［J/OL］. Engineering Mechanics, 2023: 113. 20231127. ［20240426］.http://kns.cnki.net/kcms/detail/11.2595.O3.20230612.1556.003.html.
［7］薛建阳, 路鹏, 董晓阳. 古建筑木结构歪闪斗栱抗震性能的ABAQUS有限元分析［J］. 世界地震工程, 2017,33(4):1117.
XUE Jianyang, LU Peng, DONG Xiaoyang. Finite element analysis on the seismic behavior of the skewed DOUGONG brackets in ancient timber buildings by ABAQUS program［J］. World Earthquake Engineering, 2017,33(4): 1117.
［8］薛建阳, 路鹏, 董晓阳. 古建筑木结构歪闪斗拱竖向受力性能的ABAQUS有限元分析［J］. 西安建筑科技大学学报(自然科学版), 2017,49(1): 813.
XUE Jianyang, LU Peng, DONG Xiaoyang. Finite element analysis on the vertically mechanical behavior of the skewed DouGong in ancient timber buildings by ABAQUS program［J］. J. Xi′an Univ. of Arch. & Tech. (Natural Science Edition), 2017,49(1): 813.
［9］叶肖伟, 董传智. 基于计算机视觉的结构位移监测综述［J］. 中国公路学报, 2019,32(11): 2139.
YE Xiaowei, DONG Chuanzhi. Review of computer visionbased structural displacement monitoring［J］. China Journal of Highway and Transport, 2019,32(11): 2139.
［10］ATHA D J, JAHANSHAHI M R. Evaluation of deep learning approaches based on convolutional neural networks for corrosion detection［J］. Structural Health Monitoring, 2018, 17(5): 11101128.
［11］LIU J J, LIU C Y, WU Y Q, et al. Insulators′ identification and missing defect detection in aerial images based on cascaded YOLO models［J］. Computational Intelligence and Neuroscience, 2022: 115.
［12］杨娜, 张翀, 李天昊. 基于无人机与计算机视觉的中国古建筑木结构裂缝监测系统设计 ［J］. 工程力学, 2021,38(3): 2739.
YANG Na, ZHANG Chong, LI Tianhao. Design of crack monitoring system for Chinese ancient wooden buildings based on UAV and CV［J］. Engineering Mechanics, 2021,38(3): 2739.
［13］刘纲, 顾嘉伟, 李孟珠,等. 基于数字图像相关的古建筑彩绘梁损伤识别方法［J］. 土木与环境工程学报(中英文), 2022,44(2): 148157.
LIU Gang, GU Jiawei, LI Mengzhu, et al. Damage identification method of painted beams based on digital image correlation［J］. Journal of Civil and Environmental Engineering, 2022, 44(2): 148157.
［14］马健, 阎卫东, 刘国奇. 基于YOLO v5的古建筑木结构裂缝检测方法［J］. 沈阳建筑大学学报(自然科学版), 2021,37(5): 927934.
MA Jian, YAN Weidong, LIU Guoqi. Research on crack detection method of wooden ancient building based on YOLO v5［J］. Journal of Shenyang Jianzhu University (Natural Science), 2021, 37(5): 927934.
［15］杨娜,王烁,汪德佳. 基于计算机视觉的藏式古建筑石砌体壁画墙裂缝生长变形监测［J/OL］.工程力学, 2023: 115. 20230522. ［20240426］.http://kns.cnki.net/kcms/detail/11.2595.O3.20230519.1729.016.html.
YANG Na, WANG Shuo, WANG Dejia. Deformation monitoring during the growth of cracks inmural walls with stone masonry material of Tibetan ancient buildings based on computer vision［J/OL］. Engineering Mechanics, 2023: 115. 20230522. ［20240426］. http://kns.cnki.net/kcms/dtail/11.2595.O3.20230519.1729.016.html.
［16］WANG N N, ZHAO X f, ZOU Z, et al. Autonomous damage segmentation and measurement of glazed tiles in historic buildings via deep learning［J］. ComputerAided Civil and Infrastructure Engineering, 2020, 35(3): 277291.
［17］HU D A, YU M, WU X Y, et al. DGWYOLOv8: A small insulator target detection algorithm based on deformable attention backbone and WIoU loss function［J］. IET Image Processing, 2024,18(4): 10961108.
［18］ZHANG L J, FANG J J, LIU Y X, et al. CRYOLOv8: Multiscale object detection in traffic sign images［J］. IEEE Access, 2023,12: 219228.
［19］FANG S, YANG L, TANG J, et al. Visual measurement of lateral relative displacement of wheelrail of highspeed train under earthquake［J］. Engineering Structures, 2024, 305: 117736.
［20］SUN C, GU D, LU X. Threedimensional structural displacement measurement using monocular vision and deep learning based pose estimation［J］. Mechanical Systems and Signal Processing, 2023, 190: 110141.
［21］SAMET N, AKBAS E. HPRNet: Hierarchical point regression for wholebody human pose estimation［J］. Image and Vision Computing, 2021, 115: 104285.
［22］MUKHOPADHYAY P, CHAUDHURI B B. A survey of hough transform［J］. Pattern Recognition, 2015, 48(3): 9931010.
［23］LI S, SCHIEBER H, CORELL N, et al. GBOT: Graphbased 3D object tracking for augmented realityassisted assembly guidance［C］//2024 IEEE Conference Virtual Reality and 3D User Interfaces (VR). Orlando: IEEE, 2024: 513523.


备注/Memo

备注/Memo:
收稿日期:20240501修回日期:20240727
基金项目:中央基本科研业务费(2023JBZY028)
第一作者:王娟(1982—)，女，博士，教授，主要研究方向为古建筑木结构受力性能与结构健康监测．Email: juanwang@bjtueducn
更新日期/Last Update: 2024-11-21