-
摘要: 应用改进C-V模型, 进行桥梁下部结构裂缝图像分割, 通过裂缝截取、图像填充和旋转变换精确提取裂缝宽度。对不同光照条件下拍摄的在役混凝土桥梁结构裂缝图像, 分别利用改进C-V模型算法、自适应阈值法、形态学算法、C-V模型以及Canny算法进行试验对比。分析结果表明: 改进C-V模型算法误分率和运算时间最小, 分别为3.02%与89 ms; 1 000幅桥梁结构裂缝图像试验对比显示裂缝检测准确率大于90.8%, 裂缝宽度平均误差小于0.03 mm。可见, 改进算法可有效提高检测准确率, 减少运算时间。Abstract: The crack image segmentation of bridge substructure was studied by utilizing a modified C-V model.Crack clip, image filling and rotation transformation were applied for the precise extraction of crack width.The crack images of existing concrete bridge structure were taken in different illuminations, and test results were compared by using modified C-V model algorithm, adaptive threshold algorithm, morphology algorithm, C-V model and Canny algorithm.Analysis result indicates that the misclassification rate of modified C-V model algorithm is 3.02%, the operation time is 89 ms, and the values are minimum compared with other methods.Based on the comparative test on 1 000 crack images of bridge structure, the accuracy rate of crack detection is greater than 90.8%, and the mean error of crack width is less than 0.03 mm.So the modified algorithm can effectively improve detection accuracy rate, and reduce operation time.
-
Table 1. Comparison of algorithm performances
Algorithm Adaptivethreshold Morphology C-V model Iterative Canny Proposed algorithm M/% 7.86 6.62 6.87 5.63 3.02 T/ms 568 465 599 109 89 Table 2. Calculation result of crack width
Accuracy rate/% Maximum error/mm Mean error/mm > 90.8 < 0.52 < 0.03 -
[1] DILENA M, MORASSI A. Dynamic testing of a damaged bridge[J]. Mechanical Systems and Signal Processing, 2011, 25(1): 1485-1507. https://www.sciencedirect.com/science/article/pii/S0888327011000094 [2] KIM C W, KAWATANI M, OZAKI R, et al. Recovering missing data transmitted from a wireless sensor node for vibration-based bridge health monitoring[J]. Structural Engin-eering and Mechanics, 2011, 38(4): 417-428. doi: 10.12989/sem.2011.38.4.417 [3] ABDEL-QADER I, ABUDAYYEH O, KELLY M E. Analy-sis of edge-detection techniques for crack identification in bridges[J]. Journal of Computing in Civil Engineering, 2003, 17(4): 255-263. doi: 10.1061/(ASCE)0887-3801(2003)17:4(255) [4] HUTCHINSON T C, CHEN Zhi-qiang. Improved imageanalysis for evaluating concrete damage[J]. Journal of Com-puting in Civil Engineering, 2006, 20(3): 210-216. doi: 10.1061/(ASCE)0887-3801(2006)20:3(210) [5] NAVON E, MILLER O, AVERBUCH A. Color image segmentation based on adaptive local thresholds[J]. Image and Vision Computing, 2005, 23(1): 69-85. doi: 10.1016/j.imavis.2004.05.011 [6] IYER S, SINHA S K. Segmentation of pipe images for crack detection in buried sewers[J]. Computer-Aided Civil and Infrastructure Engineering, 2006, 21(6): 395-410. doi: 10.1111/j.1467-8667.2006.00445.x [7] YU S N, JANG J H, HAN C S. Auto inspection system using a mobile robot for detecting concrete cracks in a tunnel[J]. Automation in Construction, 2007, 16(3): 255-261. doi: 10.1016/j.autcon.2006.05.003 [8] SINHA S K, FIEGUTH P W. Automated detection ofcracks in buried concrete pipe images[J]. Automation in Con-struction, 2006, 15(1): 58-72. doi: 10.1016/j.autcon.2005.02.006 [9] ZOU Qin, CAO Yu, LI Qing-quan, et al. CrackTree: auto-matic crack detection from pavement images[J]. PatternRecognition Letters, 2012, 33(3): 227-238. https://www.sciencedirect.com/science/article/pii/S0167865511003795 [10] OH J K, JANG G, OH S, et al. Bridge inspection robot sys-tem with machine vision[J]. Automation in Construction, 2009, 18(7): 929-941. doi: 10.1016/j.autcon.2009.04.003 [11] ZHU Zhen-hua, GERMAN S, BRILAKIS I. Visual retrieval of concrete crack properties for automated post-earthquake structural safety evaluation[J]. Automation in Construction, 2011, 20(7): 874-883. doi: 10.1016/j.autcon.2011.03.004 [12] YAMAGUCHI T, HASHIMOTO S. Fast crack detection method for large-size concrete surface images using percolation-based image processing[J]. Machine Vision and Applications, 2010, 21(5): 797-809. doi: 10.1007/s00138-009-0189-8 [13] YANG Ming-yu, DING Huan, ZHAO Bo, et al. Chan-Vese model image segmentation with neighborhood information[J]. Journal of Computer-Aided Design & Computer Graphics, 2001, 23(3): 413-418. https://www.researchgate.net/publication/284555111_Chan-Vese_model_image_segmentation_with_neighborhood_information [14] XIAO Chun-xia, CHU Yu, ZHANG Qing. Texture image segmentation using level set function evolved by Gaussian mixture model[J]. Chinese Journal of Computers, 2010, 33(7): 1295-1304. doi: 10.3724/SP.J.1016.2010.01295 [15] LI Chun-ming, KAO C Y, GORE J C, et al. Minimization of region-scalable fitting energy for image segmentation[J]. IEEE Transactions on Image Processing, 2008, 17(10): 1940-1949. doi: 10.1109/TIP.2008.2002304 [16] ZHANG Kai-hua, ZHANG Lei, SONG Hui-hui, et al. Activecontours with selective local or global segmentation: a newformulation and level set method[J]. Image and Vision Com-puting, 2010, 28(6): 668-676. https://www.sciencedirect.com/science/article/pii/S0262885609002303