Crack Texture Feature Identification of Fiber Reinforced Concrete Based on Deep Learning

Abstract

Structural cracks in concrete have a significant influence on structural safety, so it is necessary to detect and monitor concrete cracks. Deep learning is a powerful tool for detecting cracks in concrete structures. However, it requires a large quantity of training samples and is costly in terms of computational time. In order to solve these difficulties, a deep learning target detection framework combining texture features with concrete crack data is proposed. Texture features and pre-processed concrete data are merged to increase the number of feature channels in order to reduce the demand of training samples for the model and improve training speed. With this framework, concrete crack detection can be realized even with a limited number of samples. To accomplish this aim, self-made steel fiber reinforced concrete crack data is used for comparison between our framework and those without texture feature mergence or pre-processed concrete data. The experimental results show that the number of parameters that need to be fitted in the model training and training time can be correspondingly reduced and the detection accuracy can also be improved.

Coarse–Fine Combined Bridge Crack Detection Based on Deep Learning

“…Cracks are the most common form of bridge deterioration. During operation, bridges are susceptible to cracking caused by excessive loads, poor material quality, foundation deformation, concrete shrinkage, temperature changes and other factors [1][2][3][4]. The safe use and service life of bridges are severely compromised by cracking and can result in significant economic losses and casualties in the event of an accident [5][6][7].…”

Research on the Efficiency of Bridge Crack Detection by Coupling Deep Learning Frameworks with Convolutional Neural Networks

“…According to the 2021 statistical bulletin on the development of the transportation industry from the Ministry of Transport, there were 961,100 highway bridges nationwide, with a total length (TL) of 73,802,100 m. This represents an increase of 48,400 bridges, with a TL of 7,516,600 m, compared to the end of the previous year, of which 7417 were super-large bridges with a TL of 13,478,700 m, and 134,500 were medium and small bridges with a TL of 37,158,900 m. Therefore, it can be seen that the current number of bridges in China is quite large. During the operation of the bridges, cracks are prone to occur due to excessive loads, poor-quality construction materials, heat, and other factors [ 1 , 2 , 3 ]. Cracking is a severe problem for bridges, as it can significantly affect their safe operation, resulting in economic losses and casualties once an accident occurs [ 4 , 5 , 6 ].…”

Research on Crack Disease Identification Based on Visible Spectrum in Harsh Tunnel Environment

“…In the area of image segmentation, Kang et al [4] proposed a new deep encoder and decoder based network to detect pixel level cracks in complex scenes by improving/enhancing the dataset and performance.Ali et al [5]GAN was used to generate synthetic image data in order to multiply the dataset and to segment the internal damage of concrete components at pixel level using active thermography, but this method cannot be used when the concrete is wet or other disturbing factors.Choi et al [6] proposed an original convolutional neural network, the model consists of standard convolution, densely connected separable convolutional modules, a modified spatial pyramid module, and a decoder module, and verified to have a good performance in recognising cracks in urban streets by collecting the produced dataset.Protopapadakis et al [7] proposed a crack detection mechanism for concrete tunnel surfaces that utilises deep convolutional neural networks and domain-specific heuristic post-processing techniques for data processing and was validated at the Egnatia motorway tunnel in Metsovo, Greece.Makantasis et al [8] proposed a deep learning based approach for the detection of concrete defects in tunnels using a convolutional neural network to hierarchically construct high-level features from low-level features for describing the defects, as well as a multilayer perceptron to perform the detection task.Zhou et al [9] proposed a deep learning target detection framework combining texture features and concrete crack data by merging texture features and preprocessed concrete data to increase the number of feature channels.Zhou et al [10]Based on a deep learning approach, proposed a crack detection network consisting of a hybrid attention module based on effective embedded channel and positional information, as well as an integrated RFE and a multiscale feature fusion module for the detection of cracks on the surface of tunnel linings.…”