Convolutional neural networks for pavement roughness assessment using calibration‐free vehicle dynamics

Jeong, Jong‐Hyun; Jo, Hongki; Ditzler, Gregory

doi:10.1111/mice.12546

Cited by 45 publications

(30 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The frequency coordinate corresponding to the ith node is denoted by ω i as shown in Figure 3b, and N 2i−1 (ω) represents the shape function corresponding to its unit vertical deformation, the expression of which is shown in Equation (21). The shape function corresponding to its unit rotational angle is defined as N 2i (ω), the expression of which is shown in Equation (22).…”

Section: Updating the Estimated Frf Based On The Shape Function Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Road Roughness Estimation Based on the Vehicle Frequency Response Function

et al. 2021

View full text Add to dashboard Cite

Road roughness is an important factor in road network maintenance and ride quality. This paper proposes a road-roughness estimation method using the frequency response function (FRF) of a vehicle. First, based on the motion equation of the vehicle and the time shift property of the Fourier transform, the vehicle FRF with respect to the displacements of vehicle–road contact points, which describes the relationship between the measured response and road roughness, is deduced and simplified. The key to road roughness estimation is the vehicle FRF, which can be estimated directly using the measured response and the designed shape of the road based on the least-squares method. To eliminate the singular data in the estimated FRF, the shape function method was employed to improve the local curve of the FRF. Moreover, the road roughness can be estimated online by combining the estimated roughness in the overlapping time periods. Finally, a half-car model was used to numerically validate the proposed methods of road roughness estimation. Driving tests of a vehicle passing over a known-sized hump were designed to estimate the vehicle FRF, and the simulated vehicle accelerations were taken as the measured responses considering a 5% Gaussian white noise. Based on the directly estimated vehicle FRF and updated FRF, the road roughness estimation, which considers the influence of the sensors and quantity of measured data at different vehicle speeds, is discussed and compared. The results show that road roughness can be estimated using the proposed method with acceptable accuracy and robustness.

show abstract

Section: Updating the Estimated Frf Based On The Shape Function Methodsmentioning

confidence: 99%

“…The proposed algorithm was evaluated for different types of road roughness profiles. Jeong et al [22] proposed a deep learning estimation method utilizing the international roughness index (IRI) with the goal of using anonymous vehicles and their responses measured by a smartphone. The above methods were carried out in the time domain.…”

Section: Introductionmentioning

confidence: 99%

Road Roughness Estimation Based on the Vehicle Frequency Response Function

et al. 2021

View full text Add to dashboard Cite

show abstract

“…For pavement application, Jeong, Jo, and Ditzler (2020) utilized CNN to measure road pavement roughness based on data collected from drivers’ smartphones (i.e., vehicle accelerations).…”

Section: Methodsmentioning

confidence: 99%

Machine learning and image processing approaches for estimating concrete surface roughness using basic cameras

Valikhani

Jahromi

Pouyanfar

et al. 2020

Computer aided Civil Eng

View full text Add to dashboard Cite

Casting concrete at different ages for new construction and repairing or retrofitting concrete structures requires a sufficient bond between concrete casts. The bond strength between different casts is attributed to surface roughness. Surface roughness can be achieved in many ways, such as water‐jetting or sandblasting. To evaluate the degree of surface roughness, qualitative and quantitative methods are introduced by many researchers; however, several drawbacks are associated with most of these methods, including cost, availability, human errors, and inability to assess old structures from prior inspection records. Two novel industrial implementation methods are introduced in this paper to estimate, quantitatively, the concrete surface roughness from images with sufficient resolution. In the first application method, a digital image processing method is proposed to distinguish the coarse aggregate from cement paste, and a new index is presented as a function of aggregate proportional area to the surface area. In the second application method, data augmentation and transfer learning techniques in computer vision and machine learning are utilized to classify new images based on predefined images during the learning process. Both application methods were related to a well‐established method of 3D laser scanning from sandblasted concrete surfaces. Finally, a brand new set of images of sandblasted surfaces was used to test and validate both methods. The results show that both methods successfully estimate the concrete surface roughness with an accuracy of more than 93%.

show abstract

“…For pavement assessment and crack detection, CrackNet and CrackNet‐V for pixel‐level cracking detection on 3D asphalt images were developed (Fei et al., 2019; A. Zhang et al., 2017). Jeong, Jo, and Ditzler (2020) assessed the pavement roughness by using an optimized CNN. For concrete structure damage evaluation, there were studies on the reinforced concrete building damage detection using ResNet‐50 and ResNet‐50‐based YOLOv2 (Pan & Yang, 2020), pixel‐level multiple damage detection of concrete structure by using a fine‐tuned DesNet‐121 (S. Li, Zhao, & Zhou, 2019), and concrete crack detection by using context‐aware semantic segmentation (X. Zhang, Rajan, & Story, 2019).…”

Section: Introductionmentioning

confidence: 99%

Automatic railroad track components inspection using real‐time instance segmentation

Guo

Qian

et al. 2020

Computer aided Civil Eng

View full text Add to dashboard Cite

In the United States, to ensure railroad safety and keep its efficient operation, regular track inspections on track component defects are required by the Federal Railroad Administration (FRA). Various types of inspection equipment are applied, such as ground penetrating radar, laser, and LiDAR, but they are usually very expensive and require extensive training and rich experience to operate. To date, track inspections still rely heavily on manual inspections which are low‐efficiency, subjective, and not as accurate as desired, especially for missing and broken track components, such as spikes, clips, and tie plates. To address this issue, a real‐time pixel‐level rail components detection framework to inspect tracks timely and accurately is proposed in this study. The first public rail components image database, including rails, spikes, and clips, is built and released online. A real‐time pixel‐level detection framework with improved real‐time instance segmentation models is developed. The improved models leverage fast object detection and highly accurate instance segmentation. Backbones with more granular levels and receptive fields are implemented in the proposed models. Compared with the original YOLACT and Mask R‐CNN models, the proposed models are able to: (1) achieve 59.9 bbox mAP, and 63.6 mask mAP with the customized dataset, which are higher than the other models and (2) achieve a real‐time speed which is over 30 FPS processing a high‐resolution video (1,080 × 1,092) with a single GPU. The fast processing speed can quickly turn inspection videos into useful information to assist track maintenance. The railroad track components image dataset can be accessed at https://github.com/jonguo111/Rail_components_image_data

show abstract

Convolutional neural networks for pavement roughness assessment using calibration‐free vehicle dynamics

Cited by 45 publications

References 53 publications

Road Roughness Estimation Based on the Vehicle Frequency Response Function

Road Roughness Estimation Based on the Vehicle Frequency Response Function

Machine learning and image processing approaches for estimating concrete surface roughness using basic cameras

Automatic railroad track components inspection using real‐time instance segmentation

Contact Info

Product

Resources

About