A number of image processing techniques IPTs have been implemented for detecting civil infrastructure defects to partially replace human-conducted onsite inspections. These IPTs are primarily used to manipulate images to extract defect features, such as cracks in concrete and steel surfaces. However, the extensively varying real-world situations e.g., lighting and shadow changes can lead to challenges to the wide adoption of IPTs. To overcome these challenges, this article proposes a vision-based method using a deep architecture of convolutional neural networks CNNs for detecting concrete cracks without calculating the defect features. As CNNs are capable of learning image features automatically, the proposed method works without the conjugation of IPTs for extracting features. The designed CNN is trained on 40 K images of 256 × 256 pixel resolutions and, consequently, records with about 98% accuracy. The trained CNN is combined with a sliding window technique to scan any image size larger than 256 × 256 pixel resolutions. The robustness and adaptability of the proposed approach are tested on 55 images of 5,888 × 3,584 pixel resolutions taken from a different structure which is not used for training and validation processes under various conditions e.g., strong light spot, shadows, and very thin cracks. Comparative studies are conducted to examine the performance of the proposed CNN using traditional Canny and Sobel edge detection methods. The results show that the proposed method shows quite better performances and can indeed find concrete cracks in realistic situations.

/pdf/deep-learning-based-crack-damage-detection-using-1cucavd1rf.pdf

Deep Learning-Based Crack Damage Detection Using Convolutional Neural Networks

https://nottingham-repository.worktribe.com/preview/745918/Manuscript_KochEtAl_2015_accepted.pdf

A review on computer vision based defect detection and condition assessment of concrete and asphalt civil infrastructure

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

This article presents a new robust automated image processing method for detecting cracks in surface images of concrete structures This method involves two steps: (1) development of an image filter for detecting major cracks using genetic programming (GP), and (2) elimination of residual noise after filtering and detection of indistinct cracks by iterative applications of the image filter to the local regions surrounding the cracks The proposed method can be used for the accurate detection of cracks in surface images recorded under various conditions Moreover, the widths of the detected cracks can be quantified on the basis of the spatial derivatives of the brightness patterns The estimated crack widths are in good agreement with those measured manually The paper shows how research on structural health monitoring of bridges and structures has received significant attention in recent years

Concrete Crack Detection by Multiple Sequential Image Filtering

In this paper, we introduce a novel image-based approach to detect cracks in concrete surfaces. Crack detection is important for the inspection, diagnosis, and maintenance of concrete structures. However, conventional image-based approaches cannot achieve precise detection since the image of the concrete surface contains various types of noise due to different causes such as concrete blebs, stain, insufficient contrast, and shading. In order to detect the cracks with high fidelity, we assume that they are composed of thin interconnected textures and propose an image-based percolation model that extracts a continuous texture by referring to the connectivity of brightness and the shape of the percolated region, depending on the length criterion of the scalable local image processing techniques. Additionally, noise reduction based on the percolation model is proposed. We evaluated the validity of the proposed technique by using precision recall and receiver operating characteristic (ROC) analysis by means of some experiments with actual concrete surface images. Copyright © 2007 Institute of Electrical Engineers of Japan. Published by John Wiley & Sons, Inc.

Image-Based Crack Detection for Real Concrete Surfaces

The Tsukuba Challenge is a competition, in which autonomous mobile robots run on a route set on a public road under a real environment. Their task includes not only simple running but also finding multiple specific persons at the same time. This study proposes a method that would realize person searching. While many person-searching algorithms use a laser sensor and a camera in combination, our method only uses an omnidirectional camera. The search target is detected using a convolutional neural network (CNN) that performs a classification of the search target. Training a CNN requires a great amount of data for which pseudo images created by composition are used. Our method is implemented in an autonomous mobile robot, and its performance has been verified in the Tsukuba Challenge 2017.

Person Searching Through an Omnidirectional Camera Using CNN in the Tsukuba Challenge

Development of autonomous mobile robot “MML-05” based on i-Cart mini for Tsukuba challenge 2015

Generally, the bottom-up learning approaches, such as neural-network, to obtain the optimal controller of target task for mechanical system face a problem including huge number of trials, which require much time and give stress against the hardware. To avoid such problems, a simulator is often built and performed with a learning method. However, there are also problems that how simulator is constructed and how accurate it performs. In this paper, we are considering a construction of simulator directly from the real hardware. Afterward a constructed simulator is used for learning target task and the obtained optimal controller is applied to the real hardware. As an example, we picked up the pendulum swing-up task which was a typical nonlinear control problem. The construction of a simulator is performed by backpropagation method with neural-network and the optimal controller is obtained by reinforcement learning method. Both processes are implemented without using the real hardware after the data sampling, therefore, load against the hardware gets sufficiently smaller, and the objective controller can be obtained faster than using only the hardware. And we consider that our proposed method can be a basic learning strategy to obtain the optimal controller of mechanical systems.

Shingo Nakamura

Papers

Image-Based Crack Detection for Real Concrete Surfaces

Person Searching Through an Omnidirectional Camera Using CNN in the Tsukuba Challenge

Development of autonomous mobile robot “MML-05” based on i-Cart mini for Tsukuba challenge 2015

Hybrid Learning Strategy for Real Hardware of Swing-up Pendulum

Adaptive Modeling of Physical Systems Based on Affine Transform and its Application for Machine Learning