Automatic Damage Detection of Pavement through DarkNet Analysis of Digital, Infrared, and Multi-Spectral Dynamic Imaging Images.

It is important to maintain the safety of road driving by automatically performing a series of processes to automatically measure and repair damage to the road pavement. However, road pavements include not only damages such as longitudinal cracks, transverse cracks, alligator cracks, and potholes, but also various elements such as manholes, road marks, oil marks, shadows, and joints. Therefore, in order to separate categories that exist in various road pavements, in this paper, 13,500 digital, IR, and MSX images were collected and nine categories were automatically classified by DarkNet. The DarkNet classification accuracies of digital images, IR images, and MSX images are 97.4%, 80.1%, and 91.1%, respectively. The MSX image is a enhanced image of the IR image and showed an average of 6% lower accuracy than the digital image but an average of 11% higher accuracy than the IR image. Therefore, MSX images can play a complementary role if DarkNet classification is performed together with digital images. In this paper, a method for detecting the directionality of each crack through a two-dimensional wavelet transform is presented, and this result can contribute to future research on detecting cracks in pavements.

Crack Detection in Frozen Soils Using Infrared Thermographic Camera.

Frozen soils are encountered on construction sites in the polar regions or regions where artificial frozen ground (AFG) methods are used. Thus, efficient ways to monitor the behavior and potential failure of frozen soils are currently in demand. The advancement of thermographic technology presents an alternative solution as deformation occurring in frozen soils generate heat via inter-particle friction, and thus a subsequent increase in temperature. In this research, uniaxial compression tests were conducted on cylindrical frozen soil specimens of three types, namely clay, sand, and gravel. During the tests, surface temperature profiles of the specimens were recorded through an infrared video camera. The thermographic videos were analyzed, and subsequent results showed that temperature increases caused by frictional heat could be observed in all three frozen soil specimens. Therefore, increases in temperature can be deemed as an indicator for the potential failure of frozen soils and this method is applicable for monitoring purposes.

Automatic Pavement Defect Detection and Classification Using RGB-Thermal Images Based on Hierarchical Residual Attention Network.

A convolutional neural network based on an improved residual structure is proposed to implement a lightweight classification model for the recognition of complex pavement conditions, which uses RGB-thermal as input and embeds an attention module to adjust the spatial, as well as channel, information of the images. The best prediction accuracy of the proposed model is 98.88%, while the RGB-thermal is used as input and an attention mechanism is used. The attention mechanism increases the attention to detail of the image and regulates the use of image channels, which enhances the final performance of the model. It is also compared with state-of-the-art (SOTA) deep learning models, indicating our model has fewer parameters, shorter training time, and higher recognition accuracy compared to existing image classification models. A visualization method incorporating gradient-weighted class activation mapping (Grad-CAM) is proposed to analyze the classification results, comparing the data the model learns from the images under different input data.

Deep Learning Based Infrared Thermal Image Analysis of Complex Pavement Defect Conditions Considering Seasonal Effect.

Deep learning techniques underpinned by extensive data sources encompassing complex pavement features have proven effective in early pavement damage detection. With pavement features exhibiting temperature variation, inexpensive infra-red imaging technology in combination with deep learning techniques can detect pavement damages effectively. Previous experiments based on pavement data captured during summer sunny conditions when subjected to SA-ResNet deep learning architecture technique demonstrated 96.47% prediction accuracy. This paper has extended the same deep learning approach to a different dataset comprised of images captured during winter sunny conditions to compare the prediction accuracy, sensitivity and recall score with summer conditions. The results suggest that irrespective of the prevalent weather season, the proposed deep learning algorithm categorises pavement features around 92% accurately (95.18% in summer and 91.67% in winter conditions), suggesting the beneficial replacement of one image type with other. The data captured in sunny conditions during summer and winter show prediction accuracies of DC = 96.47% > MSX = 95.24% > IR-T = 93.83% and DC = 94.14% > MSX = 90.69% > IR-T = 90.173%, respectively. DC images demonstrated a sensitivity of 96.47% and 94.20% for summer and winter conditions, respectively, to demonstrate that reliable categorisation is possible with deep learning techniques irrespective of the weather season. However, summer conditions showing better overall prediction accuracy than winter conditions suggests that inexpensive IR-T imaging cameras with medium resolution levels can still be an economical solution, unlike expensive alternate options, but their usage has to be limited to summer sunny conditions.

Displacement Estimation Error in Laser Scanning Monitoring of Retaining Structures Considering Roughness.

Point clouds were obtained after laser scanning of the concrete panel, SMW, and sheet pile which is most widely used in retaining structures. The surface condition of the point cloud affects the displacement calculation, and hence both local roughness and global curvature of each point cloud were analyzed using the different sizes of the kernel. The curvature of the three retaining structures was also analyzed by the azimuth angle. In this paper, artificial displacements are generated for the point clouds of 100%, 80%, 60%, 40%, and 20% of the retaining structures, and displacement and analysis errors were calculated using the C2C, C2M, and M3C2 methods. C2C method is affected by the resolution of the point cloud, and the C2M method underestimates the displacement by the location of the points in the curvature of the retaining structures. M3C2 method had the lowest error, and the optimized M3C2 parameters for analyzing the displacement were presented.

Decomposition of 1,4-dioxane by advanced oxidation and biochemical process.

This study was undertaken to determine the optimal decomposition conditions when 1,4-dioxane was degraded using either the AOPs (Advanced Oxidation Processes) or the BAC-TERRA microbial complex. The advanced oxidation was operated with H2O2, in the range 4.7 to 51 mM, under 254 nm (25 W lamp) illumination, while varying the reaction parameters, such as the air flow rate and reaction time. The greatest oxidation rate (96%) of 1,4-dioxane was achieved with H2O2 concentration of 17 mM after a 2-hr reaction. As a result of this reaction, organic acid intermediates were formed, such as acetic, propionic and butyric acids. Furthermore, the study revealed that suspended particles, i.e., bio-flocs, kaolin and pozzolan, in the reaction were able to have an impact on the extent of 1,4-dioxane decomposition. The decomposition of 1,4-dioxane in the presence of bio-flocs was significantly declined due to hindered UV penetration through the solution as a result of the consistent dispersion of bio-particles. In contrast, dosing with pozzolan decomposed up to 98.8% of the 1,4-dioxane after 2 hr of reaction. Two actual wastewaters, from polyester manufacturing, containing 1,4-dioxane in the range 370 to 450 mg/L were able to be oxidized by as high as 100% within 15 min with the introduction of 100:200 (mg/L) Fe(II):H202 under UV illumination. Aerobic biological decomposition, employing BAC-TERRA, was able to remove up to 90% of 1,4-dioxane after 15 days of incubation. In the meantime, the by-products (i.e., acetic, propionic and valeric acid) generated were similar to those formed during the AOPs investigation. According to kinetic studies, both photo-decomposition and biodegradation of 1,4-dioxane followed pseudo first-order reaction kinetics, with k = 5 x 10(-4) s(-1) and 2.38 x 10(-6) s(-1), respectively. It was concluded that 1,4-dioxane could be readily degraded by both AOPs and BAC-TERRA, and that the actual polyester wastewater containing 1,4-dioxane could be successfully decomposed under the conditions of photo-Fenton oxidation.