Pavement cracks are a kind of common distress in road service time, and their length measurement is critical for pavement maintenance. The current automatic method of crack length measurement uses segmentation algorithms to obtain crack curves, which is time-consuming and complex. In this study, an effective method of crack length measurement was proposed and validated. The method consists of a detection module based on an object detection algorithm and a length calculation module. To increase the speed and accuracy of crack detection, an improved pavement crack detection algorithm BiFPN-enhanced YOLO V5 (YOLO V5-BiFPN) based on you look only once version 5 (YOLO V5) and bidirectional feature pyramid network (BiFPN) is proposed, and gamma correction was utilized to process pavement images. YOLO V5-BiFPN was tested in a real pavement image data set and achieved remarkable performance. In the length calculation module, the diagonal length of the crack bounding box output by the object detection algorithm can be defined as the crack length. To validate the measurement method, the true value of crack length was obtained from the segmentation data set by skeletonization. The error between the calculation result of the proposed method and the real value is 3.4%, and the average processing time of each image is 14.2 ms. The developed method addresses the problem of considerable time and financial cost associated with the existing crack length measurement methods.

The voids beneath cement concrete slabs are a major invisible disease, resulting in a rapid decrease in service performance in the composite pavement. Accurate voids prediction is essential for the extensive application and long-term service of composite pavement. This research provides a FEM-ANN (Finite Element Modelling-Artificial Neural Network) method to predict the voids beneath concrete slabs. These ANN models include the original back propagation (BP), the particle swarm optimisation (PSO) BP model, the genetic algorithm (GA) BP model, and the whale optimisation algorithm (WOA) BP model. The voids FEM model is established and validated by the measured data in the field, and the relative error of measured and simulated results is within 4%. The cross-validation results show that the WOA-BP model has the best prediction performance, with the highest score of 8, which refers to the overall score of the mean value and variance of these evaluation indices. Therefore, this FEM-ANN framework is an efficient method for estimating the voids beneath concrete slabs. Furthermore, it is discovered that the base modulus with the highest contribution degree of 20.34% is the most dominant factor in predicting the voids output.

Silicate-based surface treatment agent is a sustainable material to provide a sealing coat on the surface of Portland cement concrete to improve its hardness, impermeability, chemical durability, and abrasion resistance. A type of composite surface treatment agent of concrete is introduced in the paper, using potassium silicate and lithium silicate as main agents and different contents of lithium sulfate, urea, triethanolamine, organosilicon polyether surfactant, and fluorocarbon surfactant as additives. An orthogonal test was conducted to evaluate the effects of different proportions of the two main agents and five additives. The gelation time and viscosity of the agent and the rebound value and penetration depth of the tested concrete samples were tested to evaluate the performance of the designed composite material. The results showed that the two main agents had significant influences on the viscosity, while potassium silicate had the greatest influence on the rebound value. For penetration depth, urea, triethanolamine, organosilicon polyether nonionic surfactants, and fluorocarbon surfactants all had significant effects. The proper ingredients were selected based on the test results, and the optimum proportion of the designed surface treatment agent was determined based on variance analysis of the test results. Two representative commercial inorganic surface treatment agents were selected to compare with the optimum proportion of the designed material, and the effects of the material on the microstructures of the concrete specimens were analyzed by scanning electron microscope (SEM). The mechanism is that Calcium Silicate Hydrate (CSH) gel and ettringite generated can effectively improve surface properties of concrete. The results showed that the optimum proportion of the designed material can effectively block micropores in the surface layer, thus significantly decreasing the concrete porosity, which is conducive to the rapid improvement of concrete strength and compactness.

Crack evaluation is critical to the pavement condition acquisition and preservation. The current pavement crack detection method based on two-dimensional images is relatively mature, but the extraction accuracy is quite dependent on the quality of the images themselves, which is greatly affected by various conditions. To tackle these problems, this paper proposes a pavement crack extraction method based on three-dimensional laser point cloud, which jointly uses the M-estimator sample consensus (MSAC) algorithm and K-nearest neighbor (KNN) algorithm. According to the feature that most of the crack points are below the road surface, the MSAC algorithm is used to fit the pre-processed point cloud to the plane and separate the crack point cloud from the pavement texture points. Experimental results exhibit that the fitting effect is better when the maximum point-to-plane distance is 6 mm. The distance density is characterized by the nearest neighbor distance of each point calculated by KNN algorithm. Results show that it is better to use 0.95 times the average distance as the threshold to filter the crack area. The YOLOv5 deep learning model is conducted to identify the crack images, and the results indicate that the proposed method has a good performance in reproducing the morphology of cracks.

提出了一种基于目标检测与迭代阈值分割的道路标线分割算法.首先采用基于BiFormer改进的YOLOv5目标检测算法对道路标线区域进行快速定位与框选,然后运用快速迭代阈值分割对框选区域内的道路标线进行精细提取,最后对提取后的道路标线采用韦伯对比度进行人眼可视度评估.结果表明:该方法能够完成道路标线的快速准确提取,并实现对道路标线可视度的有效检测.

Automatic and rapid detection of pavement cracks is one of the important tasks for the highway department. This study proposed an improved model of you only look once (YOLO) V5 integrated with the vision transformer (ViT) that can calculate the attention weights of image regions and form a new feature map with weights. The ViT module was added to the neck of YOLO V5 to improve the speed and accuracy of the model. 1944 asphalt pavement images were collected for testing. The test results showed that the proposed model obtained high accuracy and speed for longitudinal, transverse, and fatigue cracks and was capable of real-time detection. The ViT-improved YOLO V5m obtained 0.872 in mAP(0.5), and the detection time for a single image is 11.9 ms. The study also investigated the pavement crack detection performance of the models in a rainfall environment. All models did not have satisfying detection capabilities in rainfall conditions, but the developed YOLO V5 had better performance.

With the increasing service life of highways, pavement diseases, especially asphalt road cracks are becoming more serious. To detect asphalt pavement cracks quickly and accurately, the application of deep learning has long been a question of great interest in this field. However, existing methods have disadvantages such as slow processing of images, low accuracy, and lack of evaluation of pavement conditions. In this study, a pavement damage detection and evaluation system based on deep learning is constructed, which can quickly and accurately complete the detection and segmentation of cracks and evaluate the degree of pavement damage. Based on the pre-prepared training set, a YOLOv5-based asphalt pavement crack object detection model and a U-net-based asphalt pavement crack semantic segmentation model are trained respectively, and their effectiveness is verified on the test set. In the training, the training set is preprocessed using a combination of different image enhancement methods including histogram equalization, image sharpening, and denoising, then compared by recognition results. Based on the identification results of the above two models, the area of crack is estimated and the damage degree of the pavement is evaluated by the pavement condition index from Highway Performance Assessment Standard (JTG 5210-2018). The results showed that the method performed well in crack detection and segmentation, with 80.2% mAP and 49.3% MioU, and provided a quantitative estimate of pavement condition.

The three-dimensional detection in point cloud data for pavement cracks has drawn the attention of many researchers recently. In the field of pavement surface point cloud detection, the key tasks include the identification of pavement cracks and the extraction of the location and size information of pavement cracks. Based on the point cloud data of pavement surface, we developed two methods to directly extract and detect cracks, respectively. The first method is based on the improved sliding window algorithm by combining the random sample consensus (RANSAC) technique to directly extract the crack information from point clouds. The second method is developed based on YOLOv5 to process the two-dimensional images transformed from point cloud data for automatic pavement crack detection. We also attempted to fuse the point cloud images with greyscale images as input for the YOLOv5. Analysis results show that the improved sliding window algorithm efficiently extracts pavement cracks with less noise, and the YOLOv5-based method obtains a good detection of pavement cracks. This article is part of the theme issue 'Artificial intelligence in failure analysis of transportation infrastructure and materials'.