Crack detection is important for the inspection and evaluation during the maintenance of concrete structures. However, conventional image‐based methods need extract crack features using complex image preprocessing techniques, so it can lead to challenges when concrete surface contains various types of noise due to extensively varying real‐world situations such as thin cracks, rough surface, shadows, etc. To overcome these challenges, this paper proposes an image‐based crack detection method using a deep convolutional neural network (CNN). A CNN is designed through modifying AlexNet and then trained and validated using a built database with 60000 images. Through comparing validation accuracy under different base learning rates, 0.01 was chosen as the best base learning rate with the highest validation accuracy of 99.06%, and its training result is used in the following testing process. The robustness and adaptability of the trained CNN are tested on 205 images with 3120 × 4160 pixel resolutions which were not used for training and validation. The trained CNN is integrated into a smartphone application to mobile more public to detect cracks in practice. The results confirm that the proposed method can indeed detect cracks in images from real concrete surfaces.

The detection and measurement of crack at pixel level is a challenge to existing methods. To overcome this challenge, this paper proposes a convolutional encoder-decoder network (CedNet) to detect crack from image, and the maximum widths and orientations of cracks are measured using image post-processing techniques. To realize this, a database including 1800 crack images (with 761×569 pixel resolution) taken from concrete structures is built. Then the CedNet is designed, trained and validated using the built database. The validating results show 98.90% accuracy, 93.58% precision, 94.73% recall, 93.18% F-measure, 87.23% intersection over union (IoU) of crack and 98.82% IoU of background. Subsequently, the robustness and adaptability of the trained model is tested. To measure true maximum widths and orientations of cracks, a laboratory experiment is carried out to calibrate a relation between ratio (pixel distance / real distance) and field of view (camera’s view range on concrete surface included in image) and distance from the smartphone to concrete surface. In the post-processing techniques, the perspective transformation is employed to correct distorted images caused by the existence of the oblique angles between the smartphone and concrete surfaces. Then the maximum widths and orientations of cracks in predicted results are measured respectively using the Euclidean distance transformation and least squares principle. As comparison, two existing deep learning-based crack detection and measurement method are used to examine the performance of the proposed approach. The comparison results show that the proposed method substantiates quite good performance to detect cracks and measure maximum widths and orientations of cracks in our database.

Semantic segmentation network-based methods can detect concrete damage at the pixel level. However, the performance of a single semantic segmentation network is often limited. To improve the concrete damage detection performance of a semantic segmentation network, a stacking ensemble learning-based concrete crack detection method using multiple semantic segmentation networks is proposed. To realize this method, a database including 500 images and their labels with concrete crack and spalling is built and divided into training and testing sets. At first, the training and prediction of five semantic segmentation networks (FCN-8s, SegNet, U-Net, PSPNet and DeepLabv3+) are respectively implemented on the built training set according to a five-fold cross-validation principle, where 80% of the training images are used in the training process, and 20% training images are reserved. Then, in predicting the results of reserved training images from trained semantic segmentation networks, the class labels of all pixels are collected, and then four softmax regression-based ensemble learning models are trained using the collected class labels and their true classification labels. The trained ensemble learning models are applied to regressed testing results of semantic segmentation network models. Compared with the best single semantic segmentation network, the best ensemble learning model provides performance improvement of 0.21% PA, 0.54% MPA, 3.66% MIoU, and 0.12% FWIoU, respectively. The study results show that the stacking ensemble learning strategy can indeed improve concrete damage detection performance through ensemble learning of multiple semantic segmentation networks.

The identification of loosened bolts is crucial for the early warning of structural damage and maintaining the overall stability of the structure. Most existing two-dimensional computer vision-based bolt loosening detection methods need to rely on the bolt arrangement shape for perspective transformation, to overcome this limitation, this study proposes a bolt loosening angle detection method through arrangement shape-independent perspective transformation and corner point extraction based on semantic segmentation results. First, a dataset of 1748 images containing bolts is collected. Second, a bolt region of interest (RoI) detection is developed to extract the sub-image of the bolt. The results show that the training and validation accuracy of bolt RoI detector are 99.5%, and 99.4%, respectively. Then, the contour corner points are extracted, the automatic and arrangement shape-independent perspective transformation of the image is completed using the planar homography-based image processing algorithm and the regular hexagonal characteristics of the bolt. Finally, automatic estimation of the bolt angle is performed on the corrected bolt image. The proposed method is used to detect bolt images obtained from different shooting distances, perspective angles, and lighting conditions. The results demonstrate that the method can accurately detect the looseness of bolts in bolt connections, and the quantification error is mostly less than 3°, it has the potential for real-time bolt loosening monitoring of multi-type bolted connections.

In the process of service, reinforced concrete structures have to bear both load and multi-environmental factors. The deterioration of reinforced concrete beams is critical to the durability, safety, and sustainability of reinforced concrete structures. The main aim of the present research is to determine the degradation mechanism of reinforced concrete beams subjected to sustained loading and multi-environmental factors. Reinforced concrete beam specimens were prepared, loaded and then exerted multi-environmental factors. At the end of each degradation period, the degradation of concrete (chemical contents of concrete beam surfaces, carbonation depth, compressive strength and maximum cracks) and the corrosion of steel bars (corrosion ratio and tensile strength) were continuously measured. Moreover, degraded reinforced concrete beams were flexural loaded in four-point bending failure tests. The degradation mechanism of reinforced concrete beams subjected to sustained loading and multi-environmental factors was analyzed. Thus, this study can promote a comprehensive understanding of reinforced concrete beams subjected to sustained loading and multi-environmental factors.