Autonomous concrete crack detection using deep fully convolutional neural network

Dung, Cao Vu; Anh, Le Duc

doi:10.1016/j.autcon.2018.11.028

Cited by 878 publications

(386 citation statements)

References 18 publications

Supporting

Mentioning

319

Contrasting

Unclassified

Order By: Relevance

“…Huang, Li, and Zhang (2018) suggested a deep-learning-based semantic segmentation method using a fully convolutional network (FCN) for detecting cracks and leakages in a metro shield tunnel. Dung (2019) proposed a method for concrete crack detection using a whole encoder-decoder network with the VGG16-based encoder. Zou et al (2019) proposed a method for detecting cracks in pavement and stone surface images using an encoder-decoder network.…”

Section: Deep Learning-based Crack Detectionmentioning

confidence: 99%

Encoder–decoder network for pixel‐level road crack detection in black‐box images

Bang

Park

Kim

et al. 2019

Computer aided Civil Eng

355

170

View full text Add to dashboard Cite

Timely monitoring of pavement cracks is essential for successful maintenance of road infrastructure. Accurate information concerning crack location and severity enables proactive management of the infrastructure. Black‐box cameras, which are becoming increasingly widespread at an affordable price, can be used as efficient road‐image collectors over a wide area. However, the cracks in these images are difficult to detect, because the images containing them often include objects other than roads. Thus, we propose a pixel‐level detection method for identifying road cracks in black‐box images using a deep convolutional encoder–decoder network. The encoder consists of convolutional layers of the residual network for extracting crack features, and the decoder consists of deconvolutional layers for localizing the cracks in an input image. The proposed network was trained on 427 out of 527 images extracted from black‐box videos and tested on the remaining 100 images. Compared with VGG‐16, ResNet‐50, ResNet‐101, ResNet‐200 with transfer learning, and ResNet‐152 without transfer learning, ResNet‐152 with transfer learning exhibited the best performance, achieving recall, precision, and intersection of union of 71.98%, 77.68%, and 59.65%, respectively. The experimental results prove that the proposed method is optimal for detecting cracks in black‐box images at the pixel level.

show abstract

Section: Deep Learning-based Crack Detectionmentioning

confidence: 99%

Encoder–decoder network for pixel‐level road crack detection in black‐box images

Bang

Park

Kim

et al. 2019

Computer aided Civil Eng

355

170

View full text Add to dashboard Cite

show abstract

“…The autonomous detection of concrete structure cracks can also be realised by deep learning. Dung C. V. developed a deep learning based autonomous concrete crack detection algorithm, which can reach approximately 90% of the maximum Average precision (AP) and F-Score (F1) on validation and test datasets [49]. With so many successful works in inspection tasks with deep learning, it is possible to develop autonomous corrosion and cracks detection approaches in the future.…”

Section: Inspection Technologiesmentioning

confidence: 99%

Inspection Robots in Oil and Gas Industry: a Review of Current Solutions and Future Trends

Yang

Ren

et al. 2019

2019 25th International Conference on Automation and Computing (ICAC)

View full text Add to dashboard Cite

With the increasing demands for energy, oil and gas companies have a demand to improve their efficiency, productivity and safety. Any potential corrosions and cracks on their production, storage or transportation facilities could cause disasters to both human society and the natural environment. Since many oil and gas assets are located in the extreme environment, there is an ongoing demand for robots to perform inspection tasks, which will be more cost-effective and safer. This paper provides a state of art review of inspection robots used in the oil and gas industry which including remotely operated vehicles (ROVs), autonomous underwater vehicles (AUVs), unmanned ground vehicles (UGVs) and unmanned aerial vehicles (UAVs). Different kinds of inspection robots are designed for inspecting different asset structures. The outcome of the review suggests that the reliable autonomous inspection UAVs and AUVs will gain interest among these robots and reliable autonomous localisation, environment mapping, intelligent control strategies, path planning and Non-Destructive Testing (NDT) technology will be the primary areas of research.

show abstract

“…The stages have been highlighted as concrete production at the batching plant or in-situ at the site, transportation through transit mixers, its placing using vibrators and formwork, curing of concrete, and post-construction structural health monitoring stages. Non curing (Juliafad et al, 2019) Early formwork removal (Rockstroh, 2018;Samouh et al, 2015) Plastic shrinkage (Bella et al, 2017;Ghourchian et al, 2018) Thermal cracking (Bella et al, 2017;Zhao et al, 2019) Early age cracking (Khan, Xu, Castel, & Gilbert, 2018;Safiuddin, Kaish, Woon, & Raman, 2018) Autogenous and drying shrinkage (Gilbert, Castel, Khan, South, & Mohammadi, 2018) 5 Post Construction Reinforcement and concrete corrosion (Shi, 2018;Zhou et al, 2018) Internal crack occurrence and propagation (Cai, Fu, Shang, & Shi, 2018;Dung, 2019) Moisture sensor is placed in the concrete − to monitor water content (to prevent excess addition of water during transportation), workability, and giving realtime data of the chemical characteristics of concrete.…”

Section: Problems With Current Practicesmentioning

confidence: 99%

Internet of Things (IoT) for digital concrete quality control (DCQC): A conceptual framework

Ghosh¹,

Hosseini²,

Al-Ameri³

et al. 2019

Modern Building Materials, Structures and Techniques (MBMST)

View full text Add to dashboard Cite

Concreting is generally a manual, labour intensive and time-consuming process, putting additional burden on constrained resources. Current practices of concreting are wasteful, non-sustainable and end products usually lack proper quality conformance. This paper, as the first outcome of an ongoing research project, proposes concrete as an area ripe for being disrupted by new technological developments and the wave of automation. It puts forward arguments to show that The Internet of Things (IoT), as an emerging concept, has the potential to revolutionize concreting operations, resulting in substantial time savings, confidence in its durability and enhanced quality conformance. A conceptual framework for a digital concrete quality control (DCQC) drawing upon IoT is outlined; DCQC facilitates automated lifecycle monitoring of concrete, controlled by real-time monitoring of parameters like surface humidity, temperature variance, moisture content, vibration level, and crack occurrence and propagation of concrete members through embedded sensors. Drawing upon an analytical approach, discussions provide evidence for the advantages of adopting DCQC. The proposed system is of particular appeal for practitioners, as a workable solution for reducing water, energy consumption and required man-hours for concreting procedures, as well as, providing an interface for access to real-time data, site progress monitoring, benchmarking, and predictive analytics purposes.

show abstract

Autonomous concrete crack detection using deep fully convolutional neural network

Cited by 878 publications

References 18 publications

Encoder–decoder network for pixel‐level road crack detection in black‐box images

Encoder–decoder network for pixel‐level road crack detection in black‐box images

Inspection Robots in Oil and Gas Industry: a Review of Current Solutions and Future Trends

Internet of Things (IoT) for digital concrete quality control (DCQC): A conceptual framework

Contact Info

Product

Resources

About