Automatic pavement texture recognition using lightweight few-shot learning.

Texture is a crucial characteristic of roads, closely related to their performance. The recognition of pavement texture is of great significance for road maintenance professionals to detect potential safety hazards and carry out necessary countermeasures. Although deep learning models have been applied for recognition, the scarcity of data has always been a limitation. To address this issue, this paper proposes a few-shot learning model based on the Siamese network for pavement texture recognition with a limited dataset. The model achieved 89.8% accuracy in a four-way five-shot task classifying the pavement textures of dense asphalt concrete, micro surface, open-graded friction course and stone matrix asphalt. To align with engineering practice, global average pooling (GAP) and one-dimensional convolution are implemented, creating lightweight models that save storage and training time. Comparative experiments show that the lightweight model with GAP implemented on dense layers and one-dimensional convolution on convolutional layers reduced storage volume by 94% and training time by 99%, despite a 2.9% decrease in classification accuracy. Moreover, the model with only GAP implemented on dense layers achieved the highest accuracy at 93.5%, while reducing storage volume and training time by 83% and 6%, respectively. This article is part of the theme issue 'Artificial intelligence in failure analysis of transportation infrastructure and materials'.

Adaptive carrier fringe pattern enhancement for wavelet transform profilometry through modifying intrinsic time-scale decomposition.

The uneven background illumination and random noise will degrade the quality of the optical fringe pattern, resulting in reduced accuracy or errors in phase extraction of wavelet transform profilometry (WTP). An adaptive fringe pattern enhancement method is proposed in this paper, which can effectively solve the above problems and improve the robustness of WTP. First, a modified intrinsic time-scale decomposition (MITD) algorithm is used to decompose each row of the fringe pattern adaptively, which can obtain a set of reasonable and pure proper rotation components (PRCs) with a frequency ranging from high to low and a monotonic trend. The MITD algorithm can overcome the mode mixing problem while ensuring the completeness of decomposition. Then, based on the obtained pure PRCs, an innovative background-carrier signal-noise automatic grouping strategy is proposed. Specifically, weighted-permutation entropy (WPE) is adopted to handle noise removal, and fuzzy gray correlation analysis (FGCA) is used to separate the background and carrier signal. Finally, the desired phase information can be easily and accurately extracted from the enhanced carrier signal component by a direct wavelet ridge detection method. Both the simulation and experimental results demonstrate the effectiveness and functionality of the proposed method.

The Development and Field Evaluation of an IoT System of Low-Power Vibration for Bridge Health Monitoring.

Bridge safety is important for the safety of vehicles and pedestrians. This paper presents a study on the development of a low-power wireless acceleration sensor and deployment of the sensors on a wireless gateway and cloud platform following the Internet of Things (IoT) protocols for bridge monitoring. The entire system was validated in a field test on the Chijing bridge in Shanghai. Field evaluations indicated that the developed IoT bridge monitoring system could achieve the functions of real-time data acquisition, transmission, storage and analytical processing to synthesize safety information of the bridge. The demonstrated system was promising as a complete, practical, readily available, low-cost IoT system for bridge health monitoring.

Study on Asphalt Pavement Surface Texture Degradation Using 3-D Image Processing Techniques and Entropy Theory.

Surface texture is a very important factor affecting the anti-skid performance of pavements. In this paper, entropy theory is introduced to study the decay behavior of the three-dimensional macrotexture and microtexture of road surfaces in service based on the field test data collected over more than 2 years. Entropy is found to be feasible for evaluating the three-dimensional macrotexture and microtexture of an asphalt pavement surface. The complexity of the texture increases with the increase of entropy. Under the polishing action of the vehicle load, the entropy of the surface texture decreases gradually. The three-dimensional macrotexture decay characteristics of asphalt pavement surfaces are significantly different for different mixture designs. The macrotexture decay performance of asphalt pavement can be improved by designing appropriate mixtures. Compared with the traditional macrotexture parameter Mean Texture Depth (MTD) index, entropy contains more physical information and has a better correlation with the pavement anti-skid performance index. It has significant advantages in describing the relationship between macrotexture characteristics and the anti-skid performance of asphalt pavement.

Review on Load Transfer Mechanisms of Asphalt Mixture Meso-Structure.

Asphalt mixture is a skeleton filling system consisting of aggregate and asphalt binder. Its performance is directly affected by the internal load transfer mechanism of the skeleton filling system. It is significant to understand the load transfer mechanisms for asphalt mixture design and performance evaluation. The objective of this paper is to review the research progress of the asphalt mixture load transfer mechanism. Firstly, this paper summarizes the test methods used to investigate the load transfer mechanism of asphalt mixtures. Then, an overview of the characterization of load transfer mechanism from three aspects was provided. Next, the indicators capturing contact characteristics, contact force characteristics, and force chain characteristics were compared. Finally, the load transfer mechanism of asphalt mixtures under different loading conditions was discussed. Some recommendations and conclusions in terms of load transfer mechanism characterization and evaluation were given. The related work can provide valuable references for the study of the load transfer mechanism of asphalt mixtures.

Review of the Application of Microwave Heating Technology in Asphalt Pavement Self-Healing and De-icing.

In the past decades, a large amount of research was conducted to investigate the application prospect of microwave heating technology in improving the efficiency of asphalt pavement self-healing and de-icing. This paper reviewed the achievements in this area. Firstly, the properties of asphalt concrete after microwave heating were summarized, including microwave sensitivity and heating uniformity. Then, the evaluation indicators and influence factors of the self-healing properties of the asphalt mixtures heated by microwave were reviewed. Finally, the application of microwave heating in asphalt pavement de-icing was explored. In addition, asphalt pavement aging due to microwave heating was also reviewed. It was found that microwave heating technology has good prospects in promoting asphalt pavement self-healing and de-icing. There are also some problems that should be studied in depth, such as the cost-effectiveness of microwave-sensitive additives (MSAs), the performance of the pavement with MSAs, mechanism-based self-healing performance indicators, and the aging of asphalt pavements under cycling microwave heating.

An Improved Asphalt Penetration Test Method.

A traditional penetration test only measures the total penetration within 5 s. The penetration process is not monitored, and therefore, a large amount of information on the deformation properties of asphalt is not used. This paper documents a study to use a high-speed camera to quantify the entire penetration process and use the Finite Element Method (FEM) to interpret the penetration process using a viscoelastic model. The penetration-time relationships of several asphalt binders (70#, 90#, a rubber modified binder, and a styrene-butadiene-styrene (SBS) modified binder) have been acquired using the new method, and the FEM modeling of the penetration processes is performed. The results show that both stress relaxation and creep appear during the penetration process. The results indicate that the improved test method and its data interpretation procedure may better characterize the properties of asphalt binder, which may extend the applications of the traditional penetration test.

Influences of Interface Properties on the Performance of Fiber-Reinforced Asphalt Binder.

This paper presents an experimental study about the influence of interfacial properties on the performance of fiber-reinforced asphalt. In this study, four types of fiber including one fiber-reinforced plastic (FRP), two lignin fibers, and one basalt fiber are used, and also four types of asphalt: Asphalt No. 90, asphalt No. 70, one styrene-butadiene-styrene (SBS) modified asphalt, and asphalt rubber are used. The surface energy parameters of various asphalts and fibers and the shear strength of various fiber-reinforced asphalts are measured. On the basis of these measurements, the influences of surface properties of asphalt and fiber on the performance of fiber-reinforced asphalt are analyzed. The results show that the shear strength of asphalt binder can be significantly increased by adding fibers, and the reinforcement effect is closely related to the types of asphalt and fiber. It was discovered, for the same asphalt, that the basalt fiber has the best reinforcement effect, followed by the two lignin fibers, and the FRP. For the same fiber, asphalt rubber was the most reinforced, followed by the SBS modified asphalt, asphalt No. 70 and asphalt No. 90. It was also discovered, for the same asphalt, the higher the surface energy of the fiber, the better the fiber reinforcement effect. The analysis indicates a good correlation between the work of adhesion between asphalt and fiber and the effect of fiber reinforcement. The results can be used as a basis for the selection of the proper fiber-asphalt combination to improve fiber reinforcement effects.

Investigating the Functions of Particles in Packed Aggregate Blend using a Discrete Element Method.

In asphalt mixture, aggregates account for up to 90% of the total volume and play an important role in the mechanical characteristics of asphalt mixture. The proportions of fine and coarse aggregates in gradation, as well as the function of aggregate particles, are important factors for the skeleton structure performance of asphalt mixtures. However, the existing asphalt mixture design methods are mostly based on empirical methods, where the non-uniformity and complexity of the composition of asphalt mixtures are not fully studied. In this study, the skeleton structure of aggregate mixture and function of aggregate are studied and analyzed using the Discrete Element Method (DEM). The Particle Flow 3D (PFC3D) DEM program is used to perform the numerical simulation. The average contact number and interaction forces by aggregate particles of different sizes are obtained and studied. The skeleton structure of aggregate mixture and function of aggregate particles are further analyzed from the meso-structural perspective.