Measurement Algorithm of Building Vibration Displacement Based on Image Signal Processing
-
摘要: 针对地震后高层建筑物结构损伤监测问题，该文提出一种基于方向码匹配(OCM)和边缘增强匹配(EEM)算法的微小位移测量算法。该算法先将原始图像梯度信息与像素强度融合，增强图像信息；采用相位相关法进行匹配运算，匹配速度比归一化互相关法提升了96.1%；最后使用亚像素插值法，使测量结果达到亚像素精度。实验结果表明，该文算法避免了OCM和EEM算法量化过程中图像梯度信息的损失，大大提高了模板匹配精度，匹配速度比OCM提升了43.3%，比EEM提升了19.6%。Abstract: A micro-displacement measurement algorithm is proposed based on the Orientation Code Matching (OCM) and Edge Enhanced Matching (EEM) algorithms for monitoring the structural damage of tall buildings after earthquake. The algorithm first fuses the gradient information of the original image with the pixel intensity to enhance the image information; Then the phase correlation method is used to perform the matching operation, the matching speed is 96.1% higher than the normalized cross-correlation method; Finally, the sub-pixel interpolation method is used to make the measurement achieve sub-pixel accuracy. Experimental results show that the proposed algorithm avoids the loss of image gradient information during the quantization of OCM and EEM algorithms, greatly improves the template matching accuracy, and the matching speed is 43.3% higher than OCM and 19.6% higher than EEM.
-
Key words:
- Image processing /
- Displacement measurement /
- Template matching /
- Subpixel
-
表 1 位移测量误差对比(1.0 Hz–0.1 mm)
算法 RMSE (mm) NRMSE (%) OCM 0.1569 19.1537 EEM 0.0811 13.7379 本文算法 0.0419 12.9371 ORB 0.1861 10.2422 L_SURB 0.0455 11.5654 FRIF 0.0635 13.9175 AKAZE+BRIEF 0.0486 14.0518 
下载: 导出CSV
表 2 位移测量误差对比(1.0 Hz–0.5 mm)
算法 RMSE (mm) NRMSE (%) OCM 0.1458 8.6670 EEM 0.0762 5.5342 本文算法 0.0208 2.0162 ORB 0.4813 12.9148 L_SURB 0.0645 5.2588 FRIF 0.2678 15.0890 AKAZE+BRIEF 0.0346 3.3131
下载: 导出CSV
表 3 位移测量误差对比(1.0 Hz–2.0 mm)
算法 RMSE (mm) NRMSE (%) OCM 0.1299 2.8043 EEM 0.0694 1.5857 本文算法 0.0254 0.6234 ORB 0.5069 8.4445 L_SURB 0.0913 2.2266 FRIF 0.1148 2.7637 AKAZE+BRIEF 0.0816 1.9882
下载: 导出CSV
表 4 位移测量误差对比(1.0 Hz–5.0 mm)
算法 RMSE (mm) NRMSE (%) OCM 0.1765 1.7105 EEM 0.1360 1.3456 本文算法 0.0810 0.8077 ORB 3.6496 24.3930 L_SURB 0.3319 3.2540 FRIF 1.3921 12.2768 AKAZE+BRIEF 2.9034 17.7774
下载: 导出CSV
表 5 位移测量误差对比(2.0 mm–0.5 Hz)
算法 RMSE (mm) NRMSE (%) OCM 0.1404 3.0762 EEM 0.0951 2.2242 本文算法 0.0407 1.0115 ORB 0.6016 11.2924 L_SURB 0.1460 3.5440 FRIF 0.4837 9.8989 AKAZE+BRIEF 0.0984 2.4386
下载: 导出CSV
表 6 位移测量误差对比(2.0 mm–1.0 Hz)
算法 RMSE (mm) NRMSE (%) OCM 0.1299 2.8043 EEM 0.0694 1.5857 本文算法 0.0254 0.6234 ORB 0.5069 8.4445 L_SURB 0.0913 2.2266 FRIF 0.1148 2.7637 AKAZE+BRIEF 0.0816 1.9882
下载: 导出CSV
表 7 位移测量误差对比(2.0 mm–2.0 Hz)
算法 RMSE (mm) NRMSE (%) OCM 0.1492 3.4415 EEM 0.0974 2.2766 本文算法 0.0571 1.4305 ORB 0.7533 12.7011 L_SURB 0.1102 2.7085 FRIF 0.2911 5.2361 AKAZE+BRIEF 0.3727 9.2922
下载: 导出CSV
表 8 位移测量误差对比(2.0 mm–5.0 Hz)
算法 RMSE (mm) NRMSE (%) OCM 0.1898 4.3983 EEM 0.1253 2.9423 本文算法 0.0983 2.4761 ORB 0.7458 14.4986 L_SURB 0.5127 11.6999 FRIF 0.5313 11.4219 AKAZE+BRIEF 0.4067 10.1836
下载: 导出CSV
表 9 位移测量误差对比(EI Centro)
算法 RMSE (mm) NRMSE (%) OCM 0.1488 2.2487 EEM 0.1072 1.6262 本文算法 0.0700 1.0812 ORB 0.8118 10.5354 L_SURB 0.1379 2.1263 FRIF 0.2420 3.7104 AKAZE+BRIEF 0.1320 2.0248
下载: 导出CSV
表 10 帧间运算平均时间对比(EI Centro)
算法 平均运算时间(ms) OCM 693.5835 EEM 476.2980 本文算法 378.3580 ORB 80.6894 L-SURB 62.1746 FRIF 199.6995 AKAZE+BRIEF 45.2793
下载: 导出CSV
-
FUKUDA Y, FENG M Q, and SHINOZUKA M. Cost-effective vision-based system for monitoring dynamic response of civil engineering structures[J]. Structural Control and Health Monitoring, 2010, 17(8): 918–936. doi: 10.1002/stc.360 BREUER P, CHMIELEWSKI T, GÓRSKI P, et al. Application of GPS technology to measurements of displacements of high-rise structures due to weak winds[J]. Journal of Wind Engineering and Industrial Aerodynamics, 2002, 90(3): 223–230. doi: 10.1016/S0167-6105(01)00221-5 吴元. 一种基于参数更新的机载SAR图像目标定位方法[J]. 电子与信息学报, 2019, 41(5): 1063–1068. doi: 10.11999/JEIT180564WU Yuan. An airborne SAR image target location algorithm based on parameter refining[J]. Journal of Electronics &Information Technology, 2019, 41(5): 1063–1068. doi: 10.11999/JEIT180564 RUBLEE E, RABAUD V, KONOLIGE K, et al. ORB: An efficient alternative to SIFT or SURF[C]. 2011 International Conference on Computer Vision, Barcelona, Spain, 2011: 2564–2571. doi: 10.1109/ICCV.2011.6126544. SHU Caiwei and XIAO Xuezhong. ORB-oriented mismatching feature points elimination[C]. 2018 IEEE International Conference on Progress in Informatics and Computing (PIC), Suzhou, China, 2018: 246–249. doi: 10.1109/PIC.2018.8706272. WANG Yangping, YONG Jiu, ZHU Zhengping, et al. Augmented reality tracking registration based on improved KCF tracking and ORB feature detection[C]. The 7th International Conference on Information, Communication and Networks (ICICN), Macao, China, 2019: 230–233. doi: 10.1109/ICICN.2019.8834947. WANG Zhenhua, FAN Bin, and WU Fuchao. FRIF: Fast robust invariant feature[C]. British Machine Vision Conference 2013, Bristol, UK, 2013. doi: 10.5244/C.27.16. WANG Xiangyang, WANG Chao, WANG Li, et al. A fast and high accurate image copy-move forgery detection approach[J]. Multidimensional Systems and Signal Processing, 2020, 31(3): 857–883. doi: 10.1007/s11045-019-00688-x WANG Xu, ZOU Jiabao, and SHI Daosheng. An improved ORB image feature matching algorithm based on SURF[C]. The 3rd International Conference on Robotics and Automation Engineering (ICRAE), Guangzhou, China, 2018: 218-222. doi: 10.1109/ICRAE.2018.8586755. WANG Xinzhu, LV Xuliang, LI Ling, et al. A new method of speeded up robust features image registration based on image preprocessing[C]. 2018 International Conference on Information Systems and Computer Aided Education (ICISCAE), Changchun, China, 2018: 317-321. doi: 10.1109/ICISCAE.2018.8666894. 牛燕雄, 陈梦琪, 张贺. 基于尺度不变特征变换的快速景象匹配方法[J]. 电子与信息学报, 2019, 41(3): 626–631. doi: 10.11999/JEIT180440NIU Yanxiong, CHEN Mengqi, and ZHANG He. Fast scene matching method based on scale invariant feature transform[J]. Journal of Electronics and Information Technology, 2019, 41(3): 626–631. doi: 10.11999/JEIT180440 TAREEN S A K and SALEEM Z. A comparative analysis of SIFT, SURF, KAZE, AKAZE, ORB, and BRISK[C]. 2018 International Conference on Computing, Mathematics and Engineering Technologies (iCoMET), Sukkur, Pakistan, 2018: 1–10. doi: 10.1109/ICOMET.2018.8346440. 黄建坤. 基于图像序列的桥梁形变位移测量方法[D].[硕士论文], 西南交通大学, 2018.HUANG Jiankun. Displacement measurement method for bridge deformation based on image sequence[D].[Master dissertation], Southwest Jiaotong University, 2018. FUKUDA Y, FENG M Q, NARITA Y, et al. Vision-based displacement sensor for monitoring dynamic response using robust object search algorithm[J]. IEEE Sensors Journal, 2013, 13(12): 4725–4732. doi: 10.1109/JSEN.2013.2273309 LUO Longxi and FENG M Q. Edge‐enhanced matching for gradient-based computer vision displacement measurement[J]. Computer-Aided Civil and Infrastructure Engineering, 2018, 33(12): 1019–1040. doi: 10.1111/mice.12415 刘有桥. 基于图像处理的轨道位移监测系统研究[J]. 计算机应用与软件, 2019, 36(1): 246–250, 315. doi: 10.3969/j.issn.1000-386x.2019.01.044LIU Youqiao. Track displacement monitoring system based on image processing[J]. Computer Applications and Software, 2019, 36(1): 246–250, 315. doi: 10.3969/j.issn.1000-386x.2019.01.044 LUO Longxi, FENG M Q, and WU Z Y. Robust vision sensor for multi-point displacement monitoring of bridges in the field[J]. Engineering Structures, 2018, 163: 255–266. doi: 10.1016/j.engstruct.2018.02.014 -
图(6) / 表(11)
计量
- 文章访问数: 4694
- HTML全文浏览量: 1379
- PDF下载量: 107
- 被引次数: 0
下载:
