Advances in Computer Vision-Based Civil Infrastructure Inspection and Monitoring

This article implements the state‐of‐the‐art deep learning technologies for a civil engineering application, namely recognition of structural damage from images. Inspired by ImageNet Chall...

Deep Transfer Learning for Image-Based Structural Damage Recognition

The spatial characteristics of cracks are significant indicators to assess and evaluate the health of existing buildings and infrastructures However, the current manual crack description 

Automatic Pixel-Level Crack Detection and Measurement Using Fully Convolutional Network

Comparison of deep convolutional neural networks and edge detectors for image-based crack detection in concrete

Automated EEG-based screening of depression using deep convolutional neural network.

The CrackNet, an efficient architecture based on the Convolutional Neural Network (CNN), is proposed in this article for automated pavement crack detection on 3D asphalt surfaces with explicit objective of pixel-perfect accuracy. Unlike the commonly used CNN, CrackNet does not have any pooling layers which downsize the outputs of previous layers. CrackNet fundamentally ensures pixel-perfect accuracy using the newly developed technique of invariant image width and height through all layers. CrackNet consists of five layers and includes more than one million parameters that are trained in the learning process. The input data of the CrackNet are feature maps generated by the feature extractor using the proposed line filters with various orientations, widths, and lengths. The output of CrackNet is the set of predicted class scores for all pixels. The hidden layers of CrackNet are convolutional layers and fully connected layers. CrackNet is trained with 1,800 3D pavement images and is then demonstrated to be successful in detecting cracks under various conditions using another set of 200 3D pavement images. The experiment using the 200 testing 3D images showed that CrackNet can achieve high Precision (90.13%), Recall (87.63%) and F-measure (88.86%) simultaneously. Compared with recently developed crack detection methods based on traditional machine learning and imaging algorithms, the CrackNet significantly outperforms the traditional approaches in terms of F-measure. Using parallel computing techniques, CrackNet is programmed to be efficiently used in conjunction with the data collection software.

Automated Pixel-Level Pavement Crack Detection on 3D Asphalt Surfaces Using a Deep-Learning Network

This study develops a three-dimensional (3D) Laser Railway Detection System for automated railway fastener defect detection on 3D ballastless track. The 3D laser imaging system overcomes the shortc...

Convolutional neural network for detecting railway fastener defects using a developed 3D laser system

Railway condition survey is limited by available time, particularly for high-speed railway due to the need to use its full capacity every day. Therefore, railway condition survey at high-speed at high-automation level becomes necessary. In addition, different from highway pavement condition assessment, safety is the primary factor for railway condition survey. The railway survey frequency is usually daily or weekly, while highway or runway pavement surveys are commonly conducted on annual basis. Rail profile measurement in the transverse direction, missing or broken fastener detection and slab surface crack identification are three primary railway survey items. Currently, manual or semi-automated methods are becoming feasible for railway condition surveys, but most of the applications are based on low resolution of data and have poor precision and accuracy. The repeatability and reliability of their analysis results cannot be assured. Based on the successful applications of three-dimensional (3D) laser imaging for highways and runways, the team is developing a 3D digital railway surface evaluation system. This system consists of high-performance 3D laser profilers to acquire surface details in 3D of rail tracks and supporting elements and structures at resolutions finer than 1mm. A computer vision framework is being developed to acquire and process various positioning and visual data streams in a synchronous approach. This paper discusses the status of current railway survey technologies and limitations, presents the planned implementation of new 3D laser imaging technology for high-speed and complete coverage, and possible solutions to detecting rail surface defects. Especially, fastener inspection algorithm is being developed and tested. Preliminary results reveal that the proposed 3D system under development and the fastener recognition algorithms are substantially more capable in terms of high data quality and automation level for data analysis. The objective is to initially supplement and ultimately replace traditional methods of rail safety and condition evaluation.

Xianxing Dai

Papers

Automated Pixel-Level Pavement Crack Detection on 3D Asphalt Surfaces Using a Deep-Learning Network

Convolutional neural network for detecting railway fastener defects using a developed 3D laser system

Automated Survey of Railway Conditions: A Preliminary Investigation

Railway Fastener Detection Method Based on 3D Images