Safety inspection of concrete surfaces like road and bridge surfaces is a continuous and critical task since it is closely related with structural health and reliability of such surfaces. However, it is difficult to find cracks by visual check especially for large and complex concrete surfaces like roads and bridges. Automation in structural strength monitoring of concrete surfaces has generated a lot of interest in recent years, mainly because of introduction of cheap digital cameras and microcontrollers. However, it is still tough task because of the intensity homogeneity of cracks and complexity of the background. Inspired by recent success on applying deep learning to complex computer problems like vision, object detection etc., deep learning based algorithm is proposed in this paper for detection of cracks on concrete surfaces. The proposed algorithm uses Deep Belief Network (DBN), which is trained using an image data set of 600 crack images of concrete surfaces like bridges, roads etc collected by low cost smart phones. By the analysis of experimental data, the algorithm successfully detects images with cracks of various types. The recognition rate is more than 88% compared with 70% accuracy from a typical image based approach. The results are also compared with SVM (Support Vector Machine) and traditional approaches and the recognition rate in DBN approach has been found much higher than in these approaches. This algorithm if implemented on a robotic device or simple vehicle with image acquisition capability can prove very beneficial for non-expert inspectors, enabling them to perform crack monitoring tasks efficiently.

Deep Learning, Deep Belief Networks, Restricted Boltzmann Machine

[1] H. M. La, R. S. Lim, B. B. Basily, N. Gucunski, J. Yi, A. Maher, F. A. Romero, and H. Parvardeh, “Mechatronic systems design for an autonomous robotic system for high-efficiency bridge deck inspection and evaluation,” IEEE/ASME Trans.
[2] T. Nishikawa, J. Yoshida, T. Sugiyama, and Y. Fujino, “Concrete crack detection by multiple sequential image filtering,” Comput. Aided Civil Infrastructure Eng., vol. 27, no. 1, pp. 29–47, 2012.
[3] Z. Zhu, S. German, and I. Brilakis, “Visual retrieval of concrete crack properties for automated post-earthquake structural safety evaluation,” Autom. Construction, vol. 20, no. 7, pp. 874–883, 2011. 

[4] T. Yamaguchi and S. Hashimoto, “Fast crack detection method for large-size concrete surface images using percolation based image processing,” Mach. Vision Appl., vol. 21, no. 5, pp. 797–809, 2010. 

[5] T. Yamaguchi, S. Nakamura, and S. Hashimoto, “An efficient crack detection method using percolation-based image processing,” in Proc. 3rd IEEE Conf. Ind. Electron. Appl., Jun. 2008, pp. 1875–1880. 

[6] X. Tong, J. Guo, Y. Ling, and Z. Yin, “A new image based method for concrete bridge bottom crack detection,” in Proc. Int. Conf. Image Anal. Signal Process. (IASP), Oct. 2011, pp. 568–571. 

[7] H.N. Nguyen, T.Y. Kam, and P.Y. Cheng, “A novel automatic concrete surface crack identification using isotropic undecimated wavelet transform,” in Proc. Int. Symp. Intell. Signal Process. Commun. Syst. (ISPACS), Nov. 2012, pp. 766–771. 

[8] R. Adhikari, O. Moselhi, and A. Bagchi, “Image-based retrieval of concrete crack properties for bridge inspection,” Autom. Construction, vol. 39, pp. 180–194, 2014. 

[9] R. S. Lim, H. M. La, Z. Shan, and W. Sheng, “Developing a crack inspection robot for bridge maintenance,” in Proc. IEEE Int. Conf. Robot. Autom. (ICRA), May 2011, pp. 6288–6293.
[10] Chae M., and Abraham D., (2001). ”Neuro-Fuzzy Approaches for Sanitary Sewer Pipeline Condition Assessment.” Journal of 
Computing in Civil Engineering 15, Special issue: Information technology for life-cycle infrastructure management, Pages 4–14. doi: 10.1061/(ASCE)0887-3801(2001)15:1(4).
[11] Aws Khanfar, Mohammed Abu-Khousa, and Nasser Qaddoumi, “Microwave near-field non-destructive detection and characterization of dis-bonds in concrete structures using fuzzy logic techniques,” Composite Structures, Volume 62, Issues 3–4, Pages 335-339, ISSN 0263-8223, 10.1016/j.compstruct. 2003. 09.033, 2003

[12] H. Moon, and J. Kim, “Intelligent Crack Detecting Algorithm On The Concrete Crack Image Using Neural network,” Proceedings of the 28th ISARC, Pages 1461-1467, Seoul, Korea, 2011.

[13] Y. Fujita, Y. Mitani and Y. Hamamoto, “A Method for Crack detection on a Concrete Structure,” 18th International Conference on Pattern Recognition, Volume 3, pp. 901–904, 2006. 

[14] Gajanan K. Choudhary and Sayan Dey. Crack Detection in Concrete Surfaces using Image Processing, Fuzzy Logic, and Neural Networks‖. 2012 IEEE fifth International Conference on Advanced Computational Intelligence (ICACI) October 18-20, 2012.
[15] DING Ailing, JIAO Licheng. Pavement distress recognition based on support vector machine [J]. Journal of Changan University: Natural Science Edition, 2007, 27(2):34-37. 

[16] Yan H S, Xu D. An Approach to Estimating Product Design Time Based on Fuzzy Support Vector Machine [J]. Neural Networks IEEE Transactions on, 2007, 18(3):721-731.
[17] H. Roth, L. Lu, J. Liu, J. Yao, A. Seff, C. Kevin, L. Kim, and R. Summers, “Improving computer-aided detection using convolutional neural networks and random view aggregation,” IEEE Transactions on Medical Imaging, 2015.
[18] D. Ciresan, A. Giusti, L. M Gambardella, and J. Schmidhuber, “Deep neural networks segment neuronal membranes in electron microscopy images,” in Advances in Neural Information Processing Systems, 2012, pp. 2843–2851
[19] D. C Ciresan, A. Giusti, L. M. Gambardella, and J. Schmidhuber, “Mitosis detection in breast cancer histology images with deep neural networks,” in Medical Image Computing and Computer-Assisted Intervention (MICCAI), pp. 411–418. 2013
[20] A. Krizhevsky, I. Sutskever, and G. Hinton, “Imagenet classification with deep convolutional neural networks,” in Advances in Neural information Processing Systems, 2012, pp. 1097–1105
[21] Y. Zhang, K. Sohn, R. Villegas, G. Pan, and H. Lee, “Improving object detection with deep convolutional networks via Bayesian optimization and structured prediction,” in Proceeding of IEEE Conference on Computer Vision and Pattern Recognition, 2015, pp. 249–258.
[22] J.J. Kivinen, C. K. Williams, and N. Heess, “Visual boundary prediction: A deep neural prediction network and quality dissection,” in Proceedings of International Conference on Artificial Intelligence and Statistics, 2014, pp. 512–521.
[23] S. Varadharajan, S. Jose, K. Sharma, L. Wander, and C. Mertz, “Vision for road inspection,” in Proceedings of 2014 IEEE Winter Conference on Applications of Computer Vision, 2014, pp. 115122.
[24] Y. Hu and C. Zhao, “A local binary pattern based methods for pavement crack detection,” Journal of Pattern Recognition Research, vol. 5, no. 1, pp. 140–147, 2010.
[25] C. Zhang, N. Ji, and G. Wang, Restricted Boltzmann Machines, Chinese Journal of Engineering Mathematics [J], 2015,(2):159-173. 

[26] G. Hinton, "A practical guide to training restricted boltzmann machines," Machine Learning Group, University of Toronto, Technical report, 2010.
[27] Hinton G E. Training products of experts by minimizing contrastive divergence [J]. Neural computation, 2002, 14(8): 1771-1800. 

[28] Hinton, G., Osindero, S., and Teh, Y. “A fast learning algorithm for deep belief nets,” Neural Computation, vol. 18, 2006, pp. 1527-1554.