Research on Concrete Cracks Recognition based on Dual Convolutional Neural Network

Cracks are the most common and important diseases of concrete bridges. A dual convolutional neural network (DCN) model which is composed of one convolutional neural network (CNN) model and one fully convolutional network (FCN) model is proposed to recognize the cracks in image. Firstly, the CNN model is used to identify the crack area. The interfering factors such as spot, shadow, water stain, and graffiti in the non-crack area will be excluded by CNN model. Then, the CNN results will be segmented by the FCN model with the structure of merging layer-by-layer to extract crack features such as length and width. The DCN model is trained to recognize the actual concrete bridge cracks in this paper. The recognition results show that the DCN model has a good balance between high accuracy and low noise in the process of crack recognition compared with the current image recognition method. The reliability and accuracy of recognition are both greatly improved. The DCN model is helpful for automatic identification of cracks in concrete bridges.

[1]  Lei Hang,et al.  Survey of convolutional neural network , 2016 .

[2]  Oral Büyüköztürk,et al.  Deep Learning‐Based Crack Damage Detection Using Convolutional Neural Networks , 2017, Comput. Aided Civ. Infrastructure Eng..

[3]  Xue Yadong.  AUTOMATIC RECOGNITION OF CRACKS IN TUNNEL LINING BASED ON CHARACTERISTICS OF LOCAL GRIDS IN IMAGES , 2012 .

[4]  Mohamed S Kaseko,et al.  COMPARISON OF TRADITIONAL AND NEURAL CLASSIFIERS FOR PAVEMENT-CRACK DETECTION. , 1994 .

[5]  Gerald Penn,et al.  Convolutional Neural Networks for Speech Recognition , 2014, IEEE/ACM Transactions on Audio, Speech, and Language Processing.

[6]  Trevor Darrell,et al.  Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation , 2013, 2014 IEEE Conference on Computer Vision and Pattern Recognition.

[7]  Mahmoud R. Halfawy,et al.  Integrated Vision-Based System for Automated Defect Detection in Sewer Closed Circuit Television Inspection Videos , 2015, J. Comput. Civ. Eng..

[8]  Yoshua Bengio,et al.  Deep Sparse Rectifier Neural Networks , 2011, AISTATS.

[9]  Geoffrey E. Hinton,et al.  Reducing the Dimensionality of Data with Neural Networks , 2006, Science.

[10]  S. Hashimoto,et al.  An efficient crack detection method using percolation-based image processing , 2008, 2008 3rd IEEE Conference on Industrial Electronics and Applications.

[11]  Jong Min Lee,et al.  Machine Vision System for Automatic Inspection of Bridges , 2008, 2008 Congress on Image and Signal Processing.

[12]  Geoffrey E. Hinton,et al.  Deep Learning , 2015, Nature.

[13]  Sven Behnke,et al.  Evaluation of Pooling Operations in Convolutional Architectures for Object Recognition , 2010, ICANN.

[14]  Yoshihiko Hamamoto,et al.  A Method for Crack Detection on a Concrete Structure , 2006, 18th International Conference on Pattern Recognition (ICPR'06).

[15]  Ikhlas Abdel-Qader,et al.  ANALYSIS OF EDGE-DETECTION TECHNIQUES FOR CRACK IDENTIFICATION IN BRIDGES , 2003 .

[16]  Chang-Soo Han,et al.  Auto inspection system using a mobile robot for detecting concrete cracks in a tunnel , 2007 .

[17]  Jeong Ho Lee,et al.  Bridge inspection robot system with machine vision , 2009 .