Cracks are common defects on surfaces of man-made structures such as pavements, bridges, walls of nuclear power plants, ceilings of tunnels, etc. Timely discovering and repairing of the cracks are of great significance and importance for keeping healthy infrastructures and preventing further damages. Traditionally, the cracking inspection was conducted manually which was labor-intensive, time-consuming and costly. For example, statistics from the Central Intelligence Agency show that the world’s road network length has reached 64,285,009 km, of which the United States has 6,586,610 km. It is a huge cost to maintain and upgrade such an immense road network. Thus, fully automatic crack detection has received increasing attention. With the development of artificial intelligence (AI), the deep learning technique has achieved great success and has been viewed as the most promising way for crack detection. Based on deep learning, this research has solved four important issues existing in crack-like object detection. First, the noise problem caused by the textured background is solved by using a deep classification network to remove the non-crack region before conducting crack detection. Second, the computational efficiency is highly improved. Third, the crack localization accuracy is improved. Fourth, the proposed model is very stable and can be used to deal with a wide range of crack detection tasks. In addition, this research performs a preliminary study about the future AI system, which provides a concept that has potential to realize fully automatic crack detection without human’s intervention.

Many components of existing civil infrastructure systems, such as road pavement, bridges, and buildings, are suffered from rapid aging, which require enormous nation's resources from federal and state agencies to inspect and maintain them. Crack is one of important material and structural defects, which must be inspected not only for good maintenance of civil infrastructure with a high quality of safety and serviceability, but also for the opportunity to provide early warning against failure. Conventional human visual inspection is still considered as the primary inspection method. However, it is well established that human visual inspection is subjective and often inaccurate. In order to improve current manual visual inspection for crack detection and evaluation of civil infrastructure, this study explores the application of computer vision techniques as a non-destructive evaluation and testing (NDE&T) method for automated crack detection and quantification for different civil infrastructures. In this study, computer vision-based algorithms were developed and evaluated to deal with different situations of field inspection that inspectors could face with in crack detection and quantification. The depth, the distance between camera and object, is a necessary extrinsic parameter that has to be measured to quantify crack size since other parameters, such as focal length, resolution, and camera sensor size are intrinsic, which are usually known by camera manufacturers. Thus, computer vision techniques were evaluated with different crack inspection applications with constant and variable depths. For the fixed-depth applications, computer vision techniques were applied to two field studies, including 1) automated crack detection and quantification for road pavement using the Laser Road Imaging System (LRIS), and 2) automated crack detection on bridge cables surfaces, using a cable inspection robot. For the various-depth applications, two field studies were conducted, including 3) automated crack recognition and width measurement of concrete bridges' cracks using a high-magnification telescopic lens, and 4) automated crack quantification and depth estimation using wearable glasses with stereovision cameras. From the realistic field applications of computer vision techniques, a novel self-adaptive image-processing algorithm was developed using a series of morphological transformations to connect fragmented crack pixels in digital images. The crack-defragmentation algorithm was evaluated with road pavement images. The results showed that the accuracy of automated crack detection, associated with artificial neural network classifier, was significantly improved by reducing both false positive and false negative. Using up to six crack features, including area, length, orientation, texture, intensity, and wheel-path location, crack detection accuracy was evaluated to find the optimal sets of crack features. Lab and field test results of different inspection applications show that proposed compute vision-based crack detection and quantification algorithms can detect and quantify cracks from different structures' surface and depth. Some guidelines of applying computer vision techniques are also suggested for each crack inspection application.

Master of Science / Department of Electrical and Computer Engineering / Balasubramaniam Natarajan / In this thesis, we present an automated system for detection and classification of cracks, based on the new standard proposed by `American Association of State Highway and Transportation Officials (AASHTO)'. The AASHTO standard is a draft standard, that attempts to overcome the limitations of current crack quantifying and classification methods. In the current standard, the crack classification relies heavily on the judgment of the expert. Thus the results are susceptible to human error. The effect of human error is especially severe when the amount of data collected is large. This lead to inconsistencies even if a single standard is being followed. The new AASHTO guidelines attempt to develop a method for consistent measurement of pavement condition. Gray scale images of the road are captured by an image capture vehicle and stored on a database. Through steps of thresholding, line detect and scanning, the gray scale image is converted to binary image, with 'zeros' representing cracked pixels. PCA analysis, followed by closing and filtering operation, are carried out on the gray scale image to identify cracked sub-images. The output from the filtering operation, is then replaced with its binary counterpart. In the final step the crack parameters are calculated. The region around the crack is divided into blocks of 32x32 to approximate and calculate the crack parameters with ease. The width of the crack is approximated by the average width of crack in each block. The orientation of the crack is calculated from the angle between direction of travel and the line joining the ends of the crack. Length of the crack is the displacement between the ends of the crack, and the position of the crack is calculated from the midpoint of the line joining the end points.

<p>This thesis describes new methods for automatic crack detection in pavements. Cracks in pavements can be used as an early indication for the need of reparation. </p><p>Automatic crack detection is preferable compared to manual inventory; the repeatability can be better, the inventory can be done at a higher speed and can be done without interruption of the traffic. </p><p>The automatic and semi-automatic crack detection systems that exist today use Image Analysis methods. There are today powerful methods available in the area of Computer Vision. These methods work in higher dimensions with greater complexity and generate measures of local signal properties, while Image Analyses methods for crack detection use morphological operations on binary images. </p><p>Methods for digitalizing video data on VHS-cassettes and stitching images from nearby frames have been developed. </p><p>Four methods for crack detection have been evaluated, and two of them have been used to form a crack detection and classification program implemented in the calculation program Matlab. </p><p>One image set was used during the implementation and another image set was used for validation. The crack detection system did perform correct detection on 99.2 percent when analysing the images which were used during implementation. The result of the crack detection on the validation data was not very good. When the program is being used on data from other pavements than the one used during implementation, information about the surface texture is required to calibrate the crack detection.</p>

This thesis describes new methods for automatic crack detection in pavements. Cracks in pavements can be used as an early indication for the need of reparation. Automatic crack detection is preferable compared to manual inventory; the repeatability can be better, the inventory can be done at a higher speed and can be done without interruption of the traffic. The automatic and semi-automatic crack detection systems that exist today use Image Analysis methods. There are today powerful methods available in the area of Computer Vision. These methods work in higher dimensions with greater complexity and generate measures of local signal properties, while Image Analyses methods for crack detection use morphological operations on binary images. Methods for digitalizing video data on VHS-cassettes and stitching images from nearby frames have been developed. Four methods for crack detection have been evaluated, and two of them have been used to form a crack detection and classification program implemented in the calculation program Matlab. One image set was used during the implementation and another image set was used for validation. The crack detection system did perform correct detection on 99.2 percent when analysing the images which were used during implementation. The result of the crack detection on the validation data was not very good. When the program is being used on data from other pavements than the one used during implementation, information about the surface texture is required to calibrate the crack detection.

Surface-breaking cracks pose a serious threat to the service life of concrete structures and health monitoring is presently conducted by a visual inspection method, yielding a potential risk to safety. Diffuse ultrasonic techniques have shown their potential as an ultrasonic technique for measuring crack depth in concrete and are currently under development. In this research, the finite element method (FEM) is employed to model the ultrasound diffusion in a concrete specimen. The objectives are to use the commercial finite element (FE) tool Ansys to develop the finite element model of a concrete specimen and verify the applicability of the model by comparing with an analytic solution and experiment data. Further, various crack types are analyzed with the FE model in order to gain physical insight into the interpretation of experimental measurements. The results of this research suggest that a preliminary knowledge of the cracking process is required to correctly interpret the measured impulse responses for an unknown crack geometry, as the impulse response expresses the response of the shortest path through a system of cracks between source and receiver. Moreover, the impulse response can carry some ambiguity, as certain crack types are not uniquely determined.

Crack initiation and propagation in the bladed disks of aero-engines caused by high-cycle fatigue under cyclic loads could result in the breakdown of the engines if not detected at an early stage. Although a number of fault detection methods have been reported in the literature, it still remains very challenging to develop a reliable online technique to accurately diagnose defects in bladed disks. One of the main challenges is to characterize signals contaminated by noises. These noises caused by very dynamic engine operation environment. This work presents a new technique for engine bladed disk crack detection, which utilizes advanced analysis of clearance and time-of-arrival signals acquired from blade tip sensors. This technique involves two stages of signal processing: 1) signal pre-processing for noise elimination from predetermined causes; and 2) signal post-processing for characterizing crack initiation and location. Experimental results from the spin rig test were used to validate technique predictions.

A large-area electronics (LAE) strain sensor, termed soft elastomeric capacitor (SEC), has shown great promise in fatigue crack monitoring. The SEC is capable to monitor strain changes over a large structural surface and undergo large deformations under cracking. Previous tests verified that the SEC can detect and localize fatigue cracks under low-cycle fatigue loading. In this paper, we further investigate the SEC's capability for monitoring high-cycle fatigue cracks, which are commonly seen in steel bridges. The peak-to-peak amplitude (pk-pk amplitude) of the SEC measurement is proposed as an indicator of crack growth. This technique is is robust and insensitive to long-term capacitance drift. To overcome the difficulty of identifying the pk-pk amplitude in time series due to high signal-to-noise ratio, a signal processing method is established. This method converts the measured SEC capacitance and applied load to power spectral densities (PSD) in the frequency domain, such that the pk-pk amplitudes of the measurements can be accurately extracted. Finally, the performance of this method is validated using a fatigue test of a compact steel specimen equipped with a SEC. Results show that the crack growth under high-cycle fatigue loading can be successfully monitored using the proposed signal processing method.

Crack initiation and propagation in the bladed disks of aero-engines caused by high-cycle fatigue under cyclic loads could result in the breakdown of the engines if not detected at an early stage. Although a number of fault detection methods have been reported in the literature, it still remains very challenging to develop a reliable online technique to accurately diagnose defects in bladed disks. One of the main challenges is to characterize signals contaminated by noises. These noises caused by very dynamic engine operation environment. This work presents a new technique for engine bladed disk crack detection, which utilizes advanced analysis of clearance and time-of-arrival signals acquired from blade tip sensors. This technique involves two stages of signal processing: 1) signal pre-processing for noise elimination from predetermined causes; and 2) signal post-processing for characterizing crack initiation and location. Experimental results from the spin rig test were used to validate technique predictions.

Over the last decades, numerous techniques have been proposed worldwide to assess the actual damage of critical concrete infrastructure. A method that has progressively been used in North America is a novel microscopic tool, the Damage Rating Index (DRI). This semi-quantitative petrographic tool was developed to reliably appraise both the nature and degree of damage in concrete affected by alkali-silica reaction (ASR), which may threaten the serviceability and the durability of concrete infrastructure around the world. Performing the DRI consists of counting numerous distress features (i.e. closed and open cracks in the aggregate and cement paste) encountered on the surface of polished concrete sections (lab-made specimens or cores extracted from field structures) using a stereomicroscope at 16x magnification; once recognized and counted, the distinct distress features are multiplied by weighting factors whose purpose is to balance their relative importance towards the distress mechanism under consideration (e.g., ASR). Although reliable and efficient, performing the DRI is exceptionally time-consuming, and its results are highly operator sensitive, requiring an experienced petrographer. Therefore, this study proposes using artificial intelligence (AI) through machine learning (ML) techniques to automate the DRI test protocol estimating the damage degree of concrete affected by ASR. The ML subfield known as Deep Learning (DL) was implemented to create human-like intelligence connections using a Convolutional Neural Network (CNN) algorithm, which can predict the DRI results (machine assessment) that are close to those expected (human assessment. This research is divided into two phases: 1) performing cracks recognition using sliding windows and 2) an advanced pixel recognition. In the first phase, the results displayed some inconsistencies in cracks classification; yet, for cracks identification in the cement paste, in particular, this method presented promising results. However, the advanced pixel recognition improved the drawbacks of the first phase, providing a more accurate cracks recognition and classification. The DRI number estimation was subsequently implemented into the CNN model achieving a 74.4% accuracy. Hence, the DRI automation is a revolutionary step towards a more ubiquitous use of the method since less time is required to perform the task, besides avoiding variability among petrographers and enabling non/less experienced professionals to take advantage of this powerful microscopic tool. With a more widely accessible diagnostic tool, ASR-affected critical concrete infrastructure could be more efficiently assessed, which would ultimately increase their safety.