scholarly journals Vision Measurement of Tunnel Structures with Robust Modelling and Deep Learning Algorithms

Sensors ◽  
2020 ◽  
Vol 20 (17) ◽  
pp. 4945 ◽  
Author(s):  
Xiangyang Xu ◽  
Hao Yang
Keyword(s):  
Deep Learning ◽  
Health Monitoring ◽  
Crack Detection ◽  
Three Dimensional ◽  
Image Data ◽  
Measurement Unit ◽  
Crack Identification ◽  
Monitoring Method ◽  
Vision Measurement ◽  
Robust Modelling

The health monitoring of tunnel structures is vital to the safe operation of railway transportation systems. With the increasing mileage of tunnels, regular inspection and health monitoring are urgently demanded for the tunnel structures, especially for information regarding deformation and damage. However, traditional methods of tunnel inspection are time-consuming, expensive and highly dependent on human subjectivity. In this paper, an automatic tunnel monitoring method is investigated based on image data which is collected through the moving vision measurement unit consisting of camera array. Furthermore, geometric modelling and crack inspection algorithms are proposed where a robust three-dimensional tunnel model is reconstructed utilizing a B-spline method and crack identification is conducted by means of a Mask R-CNN network. The innovation of this investigation is that we combine the robust modelling which could be applied for the deformation analysis and the crack detection where a deep learning method is employed to recognize the tunnel cracks intelligently based on image sensors. In this study, experiments were conducted on a subway tunnel structure several kilometers long, and a robust three-dimensional model is generated and the cracks are identified automatically with the image data. The superiority of this proposal is that the comprehensive information of geometry deformation and crack damage can ensure the reliability and improve the accuracy of health monitoring.

scholarly journals Deep learning-based estimation of Flory–Huggins parameter of A–B block copolymers from cross-sectional images of phase-separated structures

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Katsumi Hagita ◽  
Takeshi Aoyagi ◽  
Yuto Abe ◽  
Shinya Genda ◽  
Takashi Honda
Keyword(s):  
Deep Learning ◽  
Diblock Copolymers ◽  
Three Dimensional ◽  
Image Data ◽  
Real Space ◽  
Cross Sectional ◽  
Regression Problem ◽  
Self Consistent Field ◽  
Experimental Image ◽  
Chain Lengths

AbstractIn this study, deep learning (DL)-based estimation of the Flory–Huggins χ parameter of A-B diblock copolymers from two-dimensional cross-sectional images of three-dimensional (3D) phase-separated structures were investigated. 3D structures with random networks of phase-separated domains were generated from real-space self-consistent field simulations in the 25–40 χN range for chain lengths (N) of 20 and 40. To confirm that the prepared data can be discriminated using DL, image classification was performed using the VGG-16 network. We comprehensively investigated the performances of the learned networks in the regression problem. The generalization ability was evaluated from independent images with the unlearned χN. We found that, except for large χN values, the standard deviation values were approximately 0.1 and 0.5 for A-component fractions of 0.2 and 0.35, respectively. The images for larger χN values were more difficult to distinguish. In addition, the learning performances for the 4-class problem were comparable to those for the 8-class problem, except when the χN values were large. This information is useful for the analysis of real experimental image data, where the variation of samples is limited.

scholarly journals Quadcopter localization and health monitoring method based on multiple virtual silhouette sensor integration

2017 ◽  
Vol 13 (7) ◽  
pp. 155014771771982
Author(s):  
Jie Hou ◽  
Baolong Guo ◽  
Juanjuan Zhu ◽  
Cheng Li ◽  
Wangpeng He
Keyword(s):  
Health Monitoring ◽  
Safety Issue ◽  
Three Dimensional ◽  
General Purpose ◽  
Monitoring Methods ◽  
Monitoring Method ◽  
Extraction Algorithm ◽  
Silhouette Information ◽  
Flight Regimes ◽  
Shadow Projection

With the widespread deployment of quadcopters, the flight safety issue attracts increasingly public and academic attentions. This article presents a quadcopter flight regime extraction algorithm for quadcopter localization and health monitoring using imageries captured by general purpose monocular cameras. First, contour information is extracted from quadcopter shadows on the ground. In order to better illustrate the three-dimensional silhouette information contained in shadow contour on the ground, a virtual sensor named Shadow Projection Tunnel is designed. Then, multiple Shadow Projection Tunnels are generated according to the extracted silhouette information and corresponding light source positions. Finally, three-dimensional quadcopter positions and flight regimes are extracted based on the aggregation between multiple Shadow Projection Tunnels. The proposed method is validated to be accurate and efficient in monitoring quadcopter position and flight regimes based on the comparative analyses. In comparison with traditional quadcopter health monitoring methods, the proposed method has advantages on deployment convenience, system robustness, precision expandability, and scenario adaptability, making it an ideal solution for quadcopter monitoring in outdoor scenarios.

scholarly journals Automated Analysis of Three-dimensional CBCT Images Taken in Natural Head Position That Combines Facial Profile Processing and Multiple Deep-learning Models

2021 ◽  
Author(s):  
Janghoon Ahn ◽  
Thong Phi Nguyen ◽  
Yoon-Ji Kim ◽  
Taeyong Kim ◽  
Jonghun Yoon
Keyword(s):  
Deep Learning ◽  
Profile Analysis ◽  
Traditional Approach ◽  
Three Dimensional ◽  
Image Data ◽  
Automated Analysis ◽  
Head Position ◽  
X Rays ◽  
Learning Models ◽  
Facial Profile

Abstract Analysing cephalometric X-rays, which is mostly performed by orthodontists or dentists, is an indispensable procedure for diagnosis and treatment planning with orthodontic patients. Artificial intelligence, especially deep-learning techniques for analysing image data, shows great potential for medical and dental image analysis and diagnosis. To explore the feasibility of automating measurement of 13 geometric parameters from three-dimensional cone beam computed tomography (CBCT) images taken in a natural head position, we here describe a smart system that combines a facial profile analysis algorithm with deep-learning models. Using multiple views extracted from the CBCT data as the dataset, our proposed method partitions and detects regions of interest by extracting the facial profile and applying Mask-RCNN, a trained decentralized convolutional neural network (CNN) that positions the key parameters. All the techniques are integrated into a software application with a graphical user interface designed for user convenience. To demonstrate the system’s ability to replace human experts, we validated the performance of the proposed method by comparing it with measurements made by two orthodontists and one advanced general dentist using a commercial dental program. The time savings compared with the traditional approach was remarkable, reducing the processing time from about 30 minutes to about 30 seconds.

An inverse elastic method of crack identification based on sparse strain sensing sheet

2020 ◽  
pp. 147592172093951 ◽  
Author(s):  
Zeyu Xiong ◽  
Branko Glisic
Keyword(s):  
Finite Element ◽  
Structural Health Monitoring ◽  
Finite Element Model ◽  
Health Monitoring ◽  
Phase Field ◽  
Direct Contact ◽  
Crack Detection ◽  
Element Model ◽  
Crack Identification ◽  
Structural Health

Reliable damage detection over large areas of structures can be achieved by spatially quasi-continuous structural health monitoring enabled by two-dimensional sensing sheets. They contain dense arrays of short-gauge sensors, which increases the probability to have sensors in direct contact with damage (e.g. crack opening) and thus identify (i.e. detect, localize, and quantify) it at an early stage. This approach in damage identification is called direct sensing. Although the sensing sheet is a reliable and low-cost technology, the overall structural health monitoring system that is using it might become complex due to large number of sensors. Hence, intentional reduction in number of sensors might be desirable. In addition, malfunction of sensors can occur in real-life settings, which results in unintentional reduction in the number of functioning sensors. In both cases, reduction in the number of (functioning) sensors may lead to lack of performance of sensing sheet. Therefore, it is important to explore the performance of sparse arrays of sensors, in the cases where sensors are not necessarily in direct contact with damage (indirect sensing). The aim of this research is to create a method for optimizing the design of arrays of sensors, that is, to find the smallest number of sensors while maintaining a satisfactory reliability of crack detection and accuracy of damage localization and quantification. To achieve that goal, we first built a phase field finite element model of cracked structure verified by the analytical model to determine the crack existence (detection), and then we used the algorithm of inverse elastostatic problem combined with phase field finite element model to determine the crack length (quantification) and location (localization) by minimizing the difference between the sensor measurements and the phase field finite element model results. In addition, we experimentally validated the method by means of a reduced-scale laboratory test and assessed the accuracy and reliability of indirect sensing.

Research on Health Monitor for Three-Dimensional Braided Composite Material

2011 ◽  
Vol 48-49 ◽  
pp. 1389-1394
Author(s):  
Yi Li ◽  
Zhen Kai Wan ◽  
Jia Lu Li
Keyword(s):  
Composite Materials ◽  
Health Monitoring ◽  
Three Dimensional ◽  
Dynamic Monitoring ◽  
Monitoring Methods ◽  
Braided Composites ◽  
Monitoring Method ◽  
Braided Composite ◽  
3D Braided Composites ◽  
Health Monitor

This paper describes two methods of Fiber Bragg Grating (FBG) embedded in the three-dimensional (3D) braded composite materials and acoustic emission health monitor for 3D braided composite materials condition. The paper mainly analyses the relations of FBG change and the inner straining under the stretching on materials. It is provided that the way of signal collection and processing. The experimental results proved that FBG sensors have exceptional sense characteristic. The braided angle of composites has a little influence on FBG signal. FBG embedded in the 3D braided composite materials has not more effect mechanical property of materials. According to two health monitoring methods and the particularity of 3D braided composites, the health monitoring method based on FBG much more applies to dynamic monitoring on 3D composites special field. This research provides basis for the study and application of advanced intelligent composites.

Identification of Asynchronous Blade Vibration Parameters by Linear Regression of Blade Tip Timing Data

2018 ◽  
Vol 140 (7) ◽  
Author(s):  
Abbas Rohani Bastami ◽  
Pedram Safarpour ◽  
Arash Mikaeily ◽  
Mohammad Mohammadi
Keyword(s):  
Health Monitoring ◽  
Crack Detection ◽  
Turbine Blades ◽  
Sampling Rate ◽  
Health Indicators ◽  
Least Mean Square ◽  
Blade Vibration ◽  
Monitoring Method ◽  
Blade Tip ◽  
Vibration Parameters

Fracture of blades is usually catastrophic and creates serious damages in the turbomachines. Blades are subjected to high centrifugal force, oscillating stresses, and high temperature which makes their life limited. Therefore, blades should be checked and replaced at specified intervals or utilize a health monitoring method for them. Crack detection by nondestructive tests can only be performed during machine overhaul which is not suitable for monitoring purposes. Blade tip timing (BTT) method as a noncontact monitoring technique is spreading for health monitoring of the turbine blades. One of the main challenges of BTT method is identification of vibration parameters from one per revolution samples which is quite below Nyquist sampling rate. In this study, a new method for derivation of blade asynchronous vibration parameters from BTT data is proposed. The proposed method requires only two BTT sensors and applies least mean square algorithm to identify frequency and amplitude of blade vibration. These parameters can be further used as blade health indicators to predict defect growth in the blades. Robustness of the proposed method against measurement noise which is an important factor has been examined by numerical simulation. An experimental test was conducted on a bladed disk to show efficiency of the proposed method.

scholarly journals Multimodal 3D DenseNet for IDH Genotype Prediction in Gliomas

Genes ◽  
2018 ◽  
Vol 9 (8) ◽  
pp. 382 ◽  
Author(s):  
Sen Liang ◽  
Rongguo Zhang ◽  
Dayang Liang ◽  
Tianci Song ◽  
Tao Ai ◽  
...  
Keyword(s):  
Deep Learning ◽  
Clinical Decision Making ◽  
Three Dimensional ◽  
Area Under The Curve ◽  
Image Data ◽  
Clinical Decision ◽  
The Cancer Genome Atlas ◽  
World Health ◽  
Validation Dataset ◽  
Who Grade

Non-invasive prediction of isocitrate dehydrogenase (IDH) genotype plays an important role in tumor glioma diagnosis and prognosis. Recently, research has shown that radiology images can be a potential tool for genotype prediction, and fusion of multi-modality data by deep learning methods can further provide complementary information to enhance prediction accuracy. However, it still does not have an effective deep learning architecture to predict IDH genotype with three-dimensional (3D) multimodal medical images. In this paper, we proposed a novel multimodal 3D DenseNet (M3D-DenseNet) model to predict IDH genotypes with multimodal magnetic resonance imaging (MRI) data. To evaluate its performance, we conducted experiments on the BRATS-2017 and The Cancer Genome Atlas breast invasive carcinoma (TCGA-BRCA) dataset to get image data as input and gene mutation information as the target, respectively. We achieved 84.6% accuracy (area under the curve (AUC) = 85.7%) on the validation dataset. To evaluate its generalizability, we applied transfer learning techniques to predict World Health Organization (WHO) grade status, which also achieved a high accuracy of 91.4% (AUC = 94.8%) on validation dataset. With the properties of automatic feature extraction, and effective and high generalizability, M3D-DenseNet can serve as a useful method for other multimodal radiogenomics problems and has the potential to be applied in clinical decision making.

Deep-learning-based crack detection with applications for the structural health monitoring of gas turbines

2019 ◽  
Vol 19 (5) ◽  
pp. 1440-1452 ◽  
Author(s):  
Mahtab Mohtasham Khani ◽  
Sahand Vahidnia ◽  
Leila Ghasemzadeh ◽  
Y Eren Ozturk ◽  
Mustafa Yuvalaklioglu ◽  
...  
Keyword(s):  
Image Processing ◽  
Deep Learning ◽  
Gas Turbines ◽  
Crack Detection ◽  
Image Data ◽  
Classification Performance ◽  
Data Set ◽  
Structural Health ◽  
Repair Costs ◽  
Classical Image

Gas turbine maintenance requires consistent inspections of cracks and other structural anomalies. The inspections provide information regarding the overall condition of the structures and yield information for estimating structural health and repair costs. Various image processing techniques have been used in the past to address the problem of automated visual crack detection with varying degrees of success. In this work, we propose a novel crack detection framework that utilizes techniques from both classical image processing and deep learning methodologies. The main contribution of this work is demonstrating that applying filters to image data in the pre-processing phase can significantly boost the classification performance of a convolutional neural network–based model. The developed architecture outperforms compared works by yielding a 96.26% classification accuracy on a data set of cracked surface images collected from gas turbines.

Research on CT Scan Image of Lung Cancer Based on Deep Learning Method in Artificial Intelligence Field

2020 ◽  
Vol 10 (4) ◽  
pp. 934-939
Author(s):  
Xiaochen Yi ◽  
Zongze Sun ◽  
Baolong Yu ◽  
Munan Yang ◽  
Zhuo Zhang
Keyword(s):  
Artificial Intelligence ◽  
Lung Cancer ◽  
Deep Learning ◽  
Cancer Diagnosis ◽  
Lung Nodule ◽  
Three Dimensional ◽  
Image Data ◽  
Heat Map ◽  
Lung Cancer Diagnosis ◽  
Nodule Detection

Cancer is one of the diseases with high mortality in the 21st century, and lung cancer ranks first in all cancer morbidity and mortality. In recent years, with the rise of big data and artificial intelligence, lung cancer-assisted diagnosis based on deep learning has gradually become A popular research topic. Computer-aided lung cancer diagnosis technology is mainly the process of processing and analyzing the lung image data obtained by medical instrument imaging. The process is summarized into four steps: medical image data preprocessing, lung parenchymal segmentation, lung Nodule detection and segmentation, as well as lesion diagnosis. In order to solve the problem that the two-dimensional image model is not applicable to three-dimensional images, this paper proposes a three-dimensional convolutional neural network model suitable for lung cancer diagnosis. The model consists of two parts. The first part is a three-dimensional deep nodule detection network (FCN) model, which generates a heat map of the lung nodules. We can locate the locations of those malignant nodules through the heat map. According to the heat map generated in the first part, the second part selects those malignant nodules that are likely to be large, and then fuses the features of these selected nodules into one feature vector, showing the whole lung scan. Finally, we use this feature to classify and determine whether we have lung cancer.

Deep learning–based autonomous concrete crack evaluation through hybrid image scanning

2019 ◽  
Vol 18 (5-6) ◽  
pp. 1722-1737 ◽  
Author(s):  
Keunyoung Jang ◽  
Namgyu Kim ◽  
Yun-Kyu An
Keyword(s):  
Decision Making ◽  
Deep Learning ◽  
Large Scale ◽  
Crack Detection ◽  
Continuous Wave ◽  
Imaging System ◽  
Concrete Specimen ◽  
Crack Identification ◽  
False Alarms ◽  
Concrete Crack

This article proposes a deep learning–based autonomous concrete crack detection technique using hybrid images. The hybrid images combining vision and infrared thermography images are able to improve crack detectability while minimizing false alarms. In particular, large-scale concrete-made infrastructures such as bridge and dam can be effectively inspected by spatially scanning the unmanned vehicle–mounted hybrid imaging system including a vision camera, an infrared camera, and a continuous-wave line laser. However, the expert-dependent decision-making for crack identification which has been widely used in industrial fields is often cumbersome, time-consuming, and unreliable. As a target concrete structure gets larger, automated decision-making becomes more desirable from the practical point of view. The proposed technique is able to achieve automated crack identification and visualization by transfer learning of a well-trained deep convolutional neural network, that is, GoogLeNet, while retaining the advantages of the hybrid images. The proposed technique is experimentally validated using a lab-scale concrete specimen with cracks of various sizes. The test results reveal that macro- and microcracks are automatically visualized while minimizing false alarms.

Sign in / Sign up

Export Citation Format

Share Document