Corroded Bolt Identification Using Mask Region-Based Deep Learning Trained on Synthesized Data

Abstract

The performance of a neural network depends on the availability of datasets, and most deep learning techniques lack accuracy and generalization when they are trained using limited datasets. Using synthesized training data is one of the effective ways to overcome the above limitation. Besides, the previous corroded bolt detection method has focused on classifying only two classes, clean and fully rusted bolts, and its performance for detecting partially rusted bolts is still questionable. This study presents a deep learning method to identify corroded bolts in steel structures using a mask region-based convolutional neural network (Mask-RCNN) trained on synthesized data. The Resnet50 integrated with a feature pyramid network is used as the backbone for feature extraction in the Mask-RCNN-based corroded bolt detector. A four-step data synthesis procedure is proposed to autonomously generate the training datasets of corroded bolts with different severities. Afterwards, the proposed detector is trained by the synthesized datasets, and its robustness is demonstrated by detecting corroded bolts in a lab-scale steel structure under varying capturing distances and perspectives. The results show that the proposed method has detected corroded bolts well and identified their corrosion levels with the most desired overall accuracy rate = 96.3% for a 1.0 m capturing distance and 97.5% for a 15° perspective angle.

DSNet: A Computer Vision‐Based Detection and Corrosion Segmentation Network for Corroded Bolt Detection in Tunnel

“…Huynh et al [28] put forward a loose bolt detection method based on an R-CNN and used the Hough line transform to estimate the loosening angle. Te improved models, mask R-CNN and faster R-CNN of R-CNNs, have been extensively adopted for anomalous (e.g., corrosion, loosening, and loss) bolt detection in engineering structures [5,[29][30][31]. Nevertheless, these models are subjected to drawbacks such as high computational demands, limited generalization ability, and reduced robustness.…”

“…Besides, the robustness of the above-mentioned algorithms is contingent on the features they have been programmed to identify. Moreover, they may exhibit low performance under lighting, orientation, and viewpoint variations, limiting the applicability of automatic defect inspection in the tunnel [5]. For the deep learning-based CV, convolutional neural networks (CNNs) dominate in the feld of the CV [6] with great success in object detection [7].…”

A pipeline corrosion detecting method using percussion and residual neural network

“…Compared with the traditional CNN, ResNet18 has a shorter running time and a higher average recognition rate. Ta et al [48] used a modified ResNet50 to extract the feature maps of bolts, making it suitable for Mask-RCNN for feature learning of bolt images. Bai et al [49] used 152 layers of ResNet to process 2D images to automatically detect structural damage in extreme events.…”

Structural Condition Assessment of Steel Anchorage Using Convolutional Neural Networks and Admittance Response

“…Nevertheless, the FOS-based method is considered not cost-effective due to the necessity of capturing reflected optical waves with extremely short wavelengths using precise and expensive interrogators [18]. Computer vision and image processing are emerging as a promising and innovative solution for monitoring civil structures, demonstrating effectiveness in identifying visible issues such as cracks, spalling, and corrosion on the surfaces of structures [19][20][21]. However, this technology faces limitations in detecting invisible progressive damages such as strand relaxation in prestressed anchorages.…”