scholarly journals Structural Crack Detection and Recognition Based on Deep Learning

2021 ◽  
Vol 11 (6) ◽  
pp. 2868
Author(s):  
Cheng Yang ◽  
Jingjie Chen ◽  
Zhiyuan Li ◽  
Yi Huang
Keyword(s):  
Neural Networks ◽  
Deep Learning ◽  
Crack Detection ◽  
Structural Safety ◽  
Training Dataset ◽  
Surface Cracks ◽  
Satisfactory Accuracy ◽  
Novel Technology ◽  
Structural Cracks ◽  
Detection And Recognition

The detection and recognition of surface cracks are of great significance for structural safety. This paper is based on a deep-learning methodology to detect and recognize structural cracks. First, a training dataset of the model is built. Then, three neural networks, AlexNet, VGGNet13, and ResNet18, are employed to recognize and classify crack images. The tests indicate that the ResNet18 model generates the most satisfactory results. It is also found that the trained YOLOv3 model detects the crack area with satisfactory accuracy. This study also confirms that the proposed deep learning as a novel technology has the potential to be an efficient and robust tool for crack detection and recognition to replace traditional methods.

scholarly journals Classification & Detection of Vehicles using Deep Learning

2020 ◽  
Vol 9 (7) ◽  
pp. 181-189
Keyword(s):  
Neural Networks ◽  
Deep Learning ◽  
Convolutional Neural Networks ◽  
Transportation Systems ◽  
Coarse Grained ◽  
Vehicle Classification ◽  
Training Dataset ◽  
License Plate ◽  
Original Dataset ◽  
Detection Of Objects

The vehicle classification and detecting its license plate are important tasks in intelligent security and transportation systems. However, theexisting methods of vehicle classification and detection are highly complex which provides coarse-grained outcomesbecause of underfitting or overfitting of the model. Due toadvanced accomplishmentsof the Deep Learning, it was efficiently implemented to image classification and detection of objects. This proposed paper come up with a new approach which makes use of convolutional neural networks concept in Deep Learning.It consists of two steps: i) vehicle classification ii) vehicle license plate recognition. Numerous classicmodules of neural networks hadbeen implemented in training and testing the vehicle classification and detection of license plate model, such as CNN (convolutional neural networks), TensorFlow, and Tesseract-OCR. The suggestedtechnique candetermine the vehicle type, number plate and other alternative dataeffectively. This model provides security and log details regarding vehicles by using AI Surveillance. It guides the surveillance operators and assists human resources. With the help of the original dataset (training) and enriched dataset (testing), this customized model(algorithm) can achieve best outcomewith a standard accuracy of around 97.32% inclassification and detection of vehicles. By enlarging the quantity of the training dataset, the loss function and mislearning rate declines progressively. Therefore, this proposedmodelwhich uses DeepLearning hadbetterperformance and flexibility. When compared to outstandingtechniques in the strategicImage datasets, this deep learning modelscan gethigher competitor outcomes. Eventually, the proposed system suggests modern methods for advancementof the customized model and forecasts the progressivegrowth of deep learningperformance in the explorationof artificial intelligence (AI) &machine learning (ML) techniques.

scholarly journals Comparative Analysis on Deep Learning Approaches for Heavy-Vehicle Detection based on Data Augmentation and Transfer-Learning techniques

2021 ◽  
Vol 13 (3) ◽  
pp. 809-820
Author(s):  
V. Sowmya ◽  
R. Radha
Keyword(s):  
Neural Networks ◽  
Deep Learning ◽  
Real Time ◽  
Transfer Learning ◽  
Convolutional Neural Networks ◽  
Traffic Management ◽  
Data Augmentation ◽  
Vehicle Detection ◽  
Heavy Vehicles ◽  
Detection And Recognition

Vehicle detection and recognition require demanding advanced computational intelligence and resources in a real-time traffic surveillance system for effective traffic management of all possible contingencies. One of the focus areas of deep intelligent systems is to facilitate vehicle detection and recognition techniques for robust traffic management of heavy vehicles. The following are such sophisticated mechanisms: Support Vector Machine (SVM), Convolutional Neural Networks (CNN), Regional Convolutional Neural Networks (R-CNN), You Only Look Once (YOLO) model, etcetera. Accordingly, it is pivotal to choose the precise algorithm for vehicle detection and recognition, which also addresses the real-time environment. In this study, a comparison of deep learning algorithms, such as the Faster R-CNN, YOLOv2, YOLOv3, and YOLOv4, are focused on diverse aspects of the features. Two entities for transport heavy vehicles, the buses and trucks, constitute detection and recognition elements in this proposed work. The mechanics of data augmentation and transfer-learning is implemented in the model; to build, execute, train, and test for detection and recognition to avoid over-fitting and improve speed and accuracy. Extensive empirical evaluation is conducted on two standard datasets such as COCO and PASCAL VOC 2007. Finally, comparative results and analyses are presented based on real-time.

scholarly journals Improved Crack Detection and Recognition Based on Convolutional Neural Network

2019 ◽  
Vol 2019 ◽  
pp. 1-8 ◽  
Author(s):  
Keqin Chen ◽  
Amit Yadav ◽  
Asif Khan ◽  
Yixin Meng ◽  
Kun Zhu
Keyword(s):  
Neural Network ◽  
Machine Learning ◽  
Neural Networks ◽  
Convolutional Neural Network ◽  
Optimization Algorithm ◽  
Crack Detection ◽  
Recognition Rate ◽  
Machine Learning Methods ◽  
Long Time ◽  
Detection And Recognition

Concrete cracks are very serious and potentially dangerous. There are three obvious limitations existing in the present machine learning methods: low recognition rate, low accuracy, and long time. Improved crack detection based on convolutional neural networks can automatically detect whether an image contains cracks and mark the location of the cracks, which can greatly improve the monitoring efficiency. Experimental results show that the Adam optimization algorithm and batch normalization (BN) algorithm can make the model converge faster and achieve the maximum accuracy of 99.71%.

On the hybridization of pre-trained deep learning and differential evolution algorithms for semantic crack detection and recognition in ensemble of infrastructures

2021 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Eslam Mohammed Abdelkader
Keyword(s):  
Deep Learning ◽  
Differential Evolution ◽  
Visual Inspection ◽  
Crack Detection ◽  
Area Under The Curve ◽  
K Nearest Neighbors ◽  
Content Type ◽  
De Algorithm ◽  
Matthew’S Correlation Coefficient ◽  
Detection And Recognition

PurposeCracks on surface are often identified as one of the early indications of damage and possible future catastrophic structural failure. Thus, detection of cracks is vital for the timely inspection, health diagnosis and maintenance of infrastructures. However, conventional visual inspection-based methods are criticized for being subjective, greatly affected by inspector's expertise, labor-intensive and time-consuming.Design/methodology/approachThis paper proposes a novel self-adaptive-based method for automated and semantic crack detection and recognition in various infrastructures using computer vision technologies. The developed method is envisioned on three main models that are structured to circumvent the shortcomings of visual inspection in detection of cracks in walls, pavement and deck. The first model deploys modified visual geometry group network (VGG19) for extraction of global contextual and local deep learning features in an attempt to alleviate the drawbacks of hand-crafted features. The second model is conceptualized on the integration of K-nearest neighbors (KNN) and differential evolution (DE) algorithm for the automated optimization of its structure. The third model is designated for validating the developed method through an extensive four layers of performance evaluation and statistical comparisons.FindingsIt was observed that the developed method significantly outperformed other crack and detection models. For instance, the developed wall crack detection method accomplished overall accuracy, F-measure, Kappa coefficient, area under the curve, balanced accuracy, Matthew's correlation coefficient and Youden's index of 99.62%, 99.16%, 0.998, 0.998, 99.17%, 0.989 and 0.983, respectively.Originality/valueLiterature review lacks an efficient method which can look at crack detection and recognition of an ensemble of infrastructures. Furthermore, there is absence of systematic and detailed comparisons between crack detection and recognition models.

scholarly journals CheXLocNet: Automatic localization of pneumothorax in chest radiographs using deep convolutional neural networks

PLoS ONE ◽  
2020 ◽  
Vol 15 (11) ◽  
pp. e0242013
Author(s):  
Hongyu Wang ◽  
Hong Gu ◽  
Pan Qin ◽  
Jia Wang
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Deep Learning ◽  
Convolutional Neural Networks ◽  
Characteristic Curve ◽  
Healthcare Delivery ◽  
Chest Radiographs ◽  
Training Dataset ◽  
Deep Convolutional Neural Networks ◽  
Support Tool

Background Pneumothorax can lead to a life-threatening emergency. The experienced radiologists can offer precise diagnosis according to the chest radiographs. The localization of the pneumothorax lesions will help to quickly diagnose, which will be benefit for the patients in the underdevelopment areas lack of the experienced radiologists. In recent years, with the development of large neural network architectures and medical imaging datasets, deep learning methods have become a methodology of choice for analyzing medical images. The objective of this study was to the construct convolutional neural networks to localize the pneumothorax lesions in chest radiographs. Methods and findings We developed a convolutional neural network, called CheXLocNet, for the segmentation of pneumothorax lesions. The SIIM-ACR Pneumothorax Segmentation dataset was used to train and validate CheXLocNets. The training dataset contained 2079 radiographs with the annotated lesion areas. We trained six CheXLocNets with various hyperparameters. Another 300 annotated radiographs were used to select parameters of these CheXLocNets as the validation set. We determined the optimal parameters by the AP50 (average precision at the intersection over union (IoU) equal to 0.50), a segmentation evaluation metric used by several well-known competitions. Then CheXLocNets were evaluated by a test set (1082 normal radiographs and 290 disease radiographs), based on the classification metrics: area under the receiver operating characteristic curve (AUC), sensitivity, specificity, and positive predictive value (PPV); segmentation metrics: IoU and Dice score. For the classification, CheXLocNet with best sensitivity produced an AUC of 0.87, sensitivity of 0.78 (95% CI 0.73-0.83), and specificity of 0.78 (95% CI 0.76-0.81). CheXLocNet with best specificity produced an AUC of 0.79, sensitivity of 0.46 (95% CI 0.40-0.52), and specificity of 0.92 (95% CI 0.90-0.94). For the segmentation, CheXLocNet with best sensitivity produced an IoU of 0.69 and Dice score of 0.72. CheXLocNet with best specificity produced an IoU of 0.77 and Dice score of 0.79. We combined them to form an ensemble CheXLocNet. The ensemble CheXLocNet produced an IoU of 0.81 and Dice score of 0.82. Our CheXLocNet succeeded in automatically detecting pneumothorax lesions, without any human guidance. Conclusions In this study, we proposed a deep learning network, called, CheXLocNet, for the automatic segmentation of chest radiographs to detect pneumothorax. Our CheXLocNets generated accurate classification results and high-quality segmentation masks for the pneumothorax at the same time. This technology has the potential to improve healthcare delivery and increase access to chest radiograph expertise for the detection of diseases. Furthermore, the segmentation results can offer comprehensive geometric information of lesions, which can benefit monitoring the sequential development of lesions with high accuracy. Thus, CheXLocNets can be further extended to be a reliable clinical decision support tool. Although we used transfer learning in training CheXLocNet, the parameters of CheXLocNet was still large for the radiograph dataset. Further work is necessary to prune CheXLocNet suitable for the radiograph dataset.

scholarly journals How Convolutional Neural Networks Diagnose Plant Disease

2019 ◽  
Vol 2019 ◽  
pp. 1-14 ◽  
Author(s):  
Yosuke Toda ◽  
Fumio Okura
Keyword(s):  
Neural Networks ◽  
Deep Learning ◽  
Convolutional Neural Networks ◽  
Plant Disease ◽  
Disease Diagnosis ◽  
Black Box ◽  
Plant Diseases ◽  
Training Dataset ◽  
Great Success ◽  
Human Decision

Deep learning with convolutional neural networks (CNNs) has achieved great success in the classification of various plant diseases. However, a limited number of studies have elucidated the process of inference, leaving it as an untouchable black box. Revealing the CNN to extract the learned feature as an interpretable form not only ensures its reliability but also enables the validation of the model authenticity and the training dataset by human intervention. In this study, a variety of neuron-wise and layer-wise visualization methods were applied using a CNN, trained with a publicly available plant disease image dataset. We showed that neural networks can capture the colors and textures of lesions specific to respective diseases upon diagnosis, which resembles human decision-making. While several visualization methods were used as they are, others had to be optimized to target a specific layer that fully captures the features to generate consequential outputs. Moreover, by interpreting the generated attention maps, we identified several layers that were not contributing to inference and removed such layers inside the network, decreasing the number of parameters by 75% without affecting the classification accuracy. The results provide an impetus for the CNN black box users in the field of plant science to better understand the diagnosis process and lead to further efficient use of deep learning for plant disease diagnosis.

scholarly journals An Efficient Approach for Detecting Driver Drowsiness Based on Deep Learning

2021 ◽  
Vol 11 (18) ◽  
pp. 8441
Author(s):  
Anh-Cang Phan ◽  
Ngoc-Hoang-Quyen Nguyen  ◽  
Thanh-Ngoan Trieu ◽  
Thuong-Cang Phan
Keyword(s):  
Neural Networks ◽  
Deep Learning ◽  
Deep Neural Networks ◽  
Training Dataset ◽  
Economic Losses ◽  
Drowsy Driving ◽  
Training Time ◽  
Road Users ◽  
Fast Training ◽  
Learning Techniques

Drowsy driving is one of the common causes of road accidents resulting in injuries, even death, and significant economic losses to drivers, road users, families, and society. There have been many studies carried out in an attempt to detect drowsiness for alert systems. However, a majority of the studies focused on determining eyelid and mouth movements, which have revealed many limitations for drowsiness detection. Besides, physiological measures-based studies may not be feasible in practice because the measuring devices are often not available on vehicles and often uncomfortable for drivers. In this research, we therefore propose two efficient methods with three scenarios for doze alert systems. The former applies facial landmarks to detect blinks and yawns based on appropriate thresholds for each driver. The latter uses deep learning techniques with two adaptive deep neural networks based on MobileNet-V2 and ResNet-50V2. The second method analyzes the videos and detects driver’s activities in every frame to learn all features automatically. We leverage the advantage of the transfer learning technique to train the proposed networks on our training dataset. This solves the problem of limited training datasets, provides fast training time, and keeps the advantage of the deep neural networks. Experiments were conducted to test the effectiveness of our methods compared with other methods. Empirical results demonstrate that the proposed method using deep learning techniques can achieve a high accuracy of 97% . This study provides meaningful solutions in practice to prevent unfortunate automobile accidents caused by drowsiness.

scholarly journals Recurrent Neural Networks for Hierarchically Mapping Human-Robot Poses

2021 ◽  
Vol 3 (1) ◽  
pp. 99-120
Author(s):  
Zainab Al-Qurashi ◽  
Brian D. Ziebart
Keyword(s):  
Neural Networks ◽  
Deep Learning ◽  
Recurrent Neural Networks ◽  
Temporal Structure ◽  
Hybrid Approach ◽  
Learning System ◽  
Training Dataset ◽  
Human Hand ◽  
Hand Position ◽  
Position And Orientation

To perform many critical manipulation tasks successfully, human-robot mimicking systems should not only accurately copy the position of a human hand, but its orientation as well. Deep learning methods trained from pairs of corresponding human and robot poses offer one promising approach for constructing a human-robot mapping to accomplish this. However, ignoring the spatial and temporal structure of this mapping makes learning it less effective. We propose two different hierarchical architectures that leverage the structural and temporal human-robot mapping. We partially separate the robotic manipulator's end-effector position and orientation while considering the mutual coupling effects between them. This divides the main problem---making the robot match the human's hand position and mimic its orientation accurately along an unknown trajectory---into several sub-problems. We address these using different recurrent neural networks (RNNs) with Long-Short Term Memory (LSTM) that we combine and train hierarchically based on the coupling over the aspects of the robot that each controls. We evaluate our proposed architectures using a virtual reality system to track human table tennis motions and compare with single artificial neural network (ANN) and RNN models. We compare the benefits of using deep learning neural networks with and without our architectures and find smaller errors in position and orientation, along with increased flexibility in wrist movement are obtained by our proposed architectures. Also, we propose a hybrid approach to collect the training dataset. The hybrid training dataset is collected by two approaches when the robot mimics human motions (standard learn from demonstrator LfD) and when the human mimics robot motions (LfDr). We evaluate the hybrid training dataset and show that the performance of the machine learning system trained by the hybrid training dataset is better with less error and faster training time compared to using the collected dataset using standard LfD approach.

Sign in / Sign up

Export Citation Format

Share Document