Design of Fast Acquisition System and Analysis of Geometric Feature for Highway Tunnel Lining Cracks Based on Machine Vision

Wang, Haozheng; Wang, Qiang; Zhai, Junli; Yuan, Dongyang; Zhang, Weikang; Xie, Xuansong; Zhou, Biao; Cai, Jielong; Lei, Y.C.

doi:10.3390/app12052516

Cited by 4 publications

(2 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Chambon [12] conducted a study of road crack detection and evaluation using computer vision. Tongji University studied an MTI-100 tunnel-detection system and achieved crack detection and location [13]. Soukup [14] used convolutional neural networks to detect the surface cracks.…”

Section: Introductionmentioning

confidence: 99%

Crack Detection of Concrete Based on Improved CenterNet Model

Kang,

Zhou,

Gao

et al. 2024

Applied Sciences

View full text Add to dashboard Cite

Cracks on concrete surfaces are vital factors affecting construction safety. Accurate and efficient crack detection can prevent safety-related accidents. Using drones to photograph cracks on a concrete surface and detect them through computer vision technology has the advantages of accurate target recognition, simple practical operation, and low cost. To solve this problem, an improved CenterNet concrete crack-detection model is proposed. Firstly, a channel-space attention mechanism is added to the original model to enhance the ability of the convolution neural network to pay attention to the image. Secondly, a feature selection module is introduced to scale the feature map in the downsampling stage to a uniform size and combine it in the channel dimension. In the upsampling stage, the feature selection module adaptively selects the combined features and fuses them with the output features of the upsampling. Finally, the target size loss is optimized from a Smooth L1 Loss to IoU Loss to lessen its inability to adapt to targets of different sizes. The experimental results show that the improved CenterNet model reduces the FPS by 123.7 Hz, increases the GPU memory by 62 MB, increases the FLOPs by 3.81 times per second, and increases the AP by 15.4% compared with the original model. The GPU memory occupancy remained stable during the training process and exhibited good real-time performance and robustness.

show abstract

Section: Introductionmentioning

confidence: 99%

Crack Detection of Concrete Based on Improved CenterNet Model

Kang,

Zhou,

Gao

et al. 2024

Applied Sciences

View full text Add to dashboard Cite

show abstract

“…Machine-vision technology is a non-contact monitoring technology without additional sensors, which does not affect traffic. Many scholars recently adopted this technology to monitor infrastructure (Wang et al, 2022;Jiang et al, 2022). Mohammed (2020) developed an expert system that helps people perform a safe roadway crossing through video cameras.…”

Section: Introductionmentioning

confidence: 99%

Machine Vision Based The Spatiotemporal Information Identification of The Vehicle

Wang

2023

JJCE

View full text Add to dashboard Cite

Accurately identifying the vehicle load on the bridge plays a vital role in structural-stress analysis and safety evaluation. Also, extracting the spatiotemporal information of the vehicle’s is crucial for identifying the vehicle load. This study aimed to propose a vehicle spatiotemporal information-identification method based on machine-vision technology. First, digital video surveillance cameras were installed in the front and on the side of the monitoring section to capture real-time videos of vehicles passing through the monitoring section. The background-difference method was used to detect vehicles based on the frontal video. Subsequently, the transverse position was evaluated according to the distance between the vehicle’s license plate and the lane line. Other vehicle parameters, including the vehicle’s speed, the number of axles and the wheelbase, were identified based on the lateral video and the auxiliary lines with a known distance. Second, a laboratory model experiment and multiple field tests under different scenes were carried out to validate the efficiency and accuracy of the proposed method. The results indicated that the average identification errors of wheelbase for the model experiment and the field tests were all 1.12% and those of the vehicle’s speed were 1.25% and 1.35, respectively. Also, the average deviations of the lateral position were 2.57 mm and 2.69 cm, respectively. The variances of the identified error of the three parameters for the field tests were 0.78%, 1.83 cm and 0.54%, respectively. This verified that the proposed method has high accuracy, reliability and good anti-noise performance. KEYWORDS: Machine vision, Spatiotemporal information, Load identification, Orthotropic deck, Bridge engineering.

show abstract

Research on the Identification and Positioning Method of Jacquard Machine Parts Based on OpenCV

李

2023

DSC

View full text Add to dashboard Cite

Design of Fast Acquisition System and Analysis of Geometric Feature for Highway Tunnel Lining Cracks Based on Machine Vision

Cited by 4 publications

References 19 publications

Crack Detection of Concrete Based on Improved CenterNet Model

Crack Detection of Concrete Based on Improved CenterNet Model

Machine Vision Based The Spatiotemporal Information Identification of The Vehicle

Research on the Identification and Positioning Method of Jacquard Machine Parts Based on OpenCV

Contact Info

Product

Resources

About