scholarly journals Wavelet Scattering and Neural Networks for Railhead Defect Identification

Materials ◽  
2021 ◽  
Vol 14 (8) ◽  
pp. 1957
Author(s):  
Yang Jin
Keyword(s):  
Neural Networks ◽  
Visual Images ◽  
Support Vector ◽  
Type I ◽  
Deep Convolutional Neural Networks ◽  
Railway Systems ◽  
Learning Framework ◽  
Signal Features ◽  
Heavy Haul ◽  
Data Requirements

Accurate and automatic railhead inspection is crucial for the operational safety of railway systems. Deep learning on visual images is effective in the automatic detection of railhead defects, but either intensive data requirements or ignoring defect sizes reduce its applicability. This paper developed a machine learning framework based on wavelet scattering networks (WSNs) and neural networks (NNs) for identifying railhead defects. WSNs are functionally equivalent to deep convolutional neural networks while containing no parameters, thus suitable for non-intensive datasets. NNs can restore location and size information. The publicly available rail surface discrete defects (RSDD) datasets were analyzed, including 67 Type-I railhead images acquired from express tracks and 128 Type-II images captured from ordinary/heavy haul tracks. The ultimate validation accuracy reached 99.80% and 99.44%, respectively. WSNs can extract implicit signal features, and the support vector machine classifier can improve the learning accuracy of NNs by over 6%. Three criteria, namely the precision, recall, and F-measure, were calculated for comparison with the literature. At the pixel level, the developed approach achieved three criteria of around 90%, outperforming former methods. At the defect level, the recall rates reached 100%, indicating all labeled defects were identified. The precision rates were around 75%, affected by the insignificant misidentified speckles (smaller than 20 pixels). Nonetheless, the developed learning framework was effective in identifying railhead defects.

Entropy estimation of the fragments of chest X-ray images

2021 ◽  
pp. 20-26
Author(s):  
Aleksei Aleksandrovich Rumyantsev ◽  
Farkhad Mansurovich Bikmuratov ◽  
Nikolai Pavlovich Pashin
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Binary Classification ◽  
Statistical Characteristics ◽  
Type I ◽  
Deep Convolutional Neural Networks ◽  
Preliminary Calculation ◽  
X Ray ◽  
The Matrix ◽  
Chest X Ray

The subject of this research is medical chest X-ray images. After fundamental pre-processing, the accumulated database of such images can be used for training deep convolutional neural networks that have become one of the most significant innovations in recent years. The trained network carries out preliminary binary classification of the incoming images and serve as an assistant to the radiotherapist. For this purpose, it is necessary to train the neural network to carefully minimize type I and type II errors. Possible approach towards improving the effectiveness of application of neural networks, by the criteria of reducing computational complexity and quality of image classification, is the auxiliary approaches: image pre-processing and preliminary calculation of entropy of the fragments. The article provides the algorithm for X-ray image pre-processing, its fragmentation, and calculation of the entropy of separate fragments. In the course of pre-processing, the region of lungs and spine is selected, which comprises approximately 30-40% of the entire image. Then the image is divided into the matrix of fragments, calculating the entropy of separate fragments in accordance with Shannon’s formula based pm the analysis of individual pixels. Determination of the rate of occurrence of each of the 255 colors allows calculating the total entropy. The use of entropy for detecting pathologies is based on the assumption that its values differ for separate fragments and overall picture of its distribution between the images with the norm and pathologies. The article analyzes the statistical values: standard deviation of error, dispersion. A fully connected neural network is used for determining the patterns in distribution of entropy and its statistical characteristics on various fragments of the chest X-ray image.

scholarly journals Deep Learning for Optic Disc Segmentation and Glaucoma Diagnosis on Retinal Images

2020 ◽  
Vol 10 (14) ◽  
pp. 4916
Author(s):  
Syna Sreng ◽  
Noppadol Maneerat ◽  
Kazuhiko Hamamoto ◽  
Khin Yadanar Win
Keyword(s):  
Neural Networks ◽  
Optic Disc ◽  
Convolutional Neural Networks ◽  
Support Vector ◽  
Retinal Images ◽  
Subjective Time ◽  
Deep Convolutional Neural Networks ◽  
Glaucoma Screening ◽  
Glaucoma Diagnosis ◽  
Eye Screening

Glaucoma is a major global cause of blindness. As the symptoms of glaucoma appear, when the disease reaches an advanced stage, proper screening of glaucoma in the early stages is challenging. Therefore, regular glaucoma screening is essential and recommended. However, eye screening is currently subjective, time-consuming and labor-intensive and there are insufficient eye specialists available. We present an automatic two-stage glaucoma screening system to reduce the workload of ophthalmologists. The system first segmented the optic disc region using a DeepLabv3+ architecture but substituted the encoder module with multiple deep convolutional neural networks. For the classification stage, we used pretrained deep convolutional neural networks for three proposals (1) transfer learning and (2) learning the feature descriptors using support vector machine and (3) building ensemble of methods in (1) and (2). We evaluated our methods on five available datasets containing 2787 retinal images and found that the best option for optic disc segmentation is a combination of DeepLabv3+ and MobileNet. For glaucoma classification, an ensemble of methods performed better than the conventional methods for RIM-ONE, ORIGA, DRISHTI-GS1 and ACRIMA datasets with the accuracy of 97.37%, 90.00%, 86.84% and 99.53% and Area Under Curve (AUC) of 100%, 92.06%, 91.67% and 99.98%, respectively, and performed comparably with CUHKMED, the top team in REFUGE challenge, using REFUGE dataset with an accuracy of 95.59% and AUC of 95.10%.

scholarly journals Deep Convolutional Neural Networks for Hyperspectral Image Classification

2015 ◽  
Vol 2015 ◽  
pp. 1-12 ◽  
Author(s):  
Wei Hu ◽  
Yangyu Huang ◽  
Li Wei ◽  
Fan Zhang ◽  
Hengchao Li
Keyword(s):  
Neural Networks ◽  
Convolutional Neural Networks ◽  
Hyperspectral Image ◽  
Image Data ◽  
Classification Performance ◽  
Spectral Signature ◽  
Support Vector ◽  
Data Sets ◽  
Deep Convolutional Neural Networks ◽  
Visual Tasks

Recently, convolutional neural networks have demonstrated excellent performance on various visual tasks, including the classification of common two-dimensional images. In this paper, deep convolutional neural networks are employed to classify hyperspectral images directly in spectral domain. More specifically, the architecture of the proposed classifier contains five layers with weights which are the input layer, the convolutional layer, the max pooling layer, the full connection layer, and the output layer. These five layers are implemented on each spectral signature to discriminate against others. Experimental results based on several hyperspectral image data sets demonstrate that the proposed method can achieve better classification performance than some traditional methods, such as support vector machines and the conventional deep learning-based methods.

scholarly journals Very Deep Convolutional Neural Networks for Complex Land Cover Mapping Using Multispectral Remote Sensing Imagery

2018 ◽  
Vol 10 (7) ◽  
pp. 1119 ◽  
Author(s):  
Masoud Mahdianpari ◽  
Bahram Salehi ◽  
Mohammad Rezaee ◽  
Fariba Mohammadimanesh ◽  
Yun Zhang
Keyword(s):  
Remote Sensing ◽  
Neural Networks ◽  
Computer Vision ◽  
Support Vector Machine ◽  
State Of The Art ◽  
Support Vector ◽  
Learning Tools ◽  
Deep Convolutional Neural Networks ◽  
Multispectral Remote Sensing

Despite recent advances of deep Convolutional Neural Networks (CNNs) in various computer vision tasks, their potential for classification of multispectral remote sensing images has not been thoroughly explored. In particular, the applications of deep CNNs using optical remote sensing data have focused on the classification of very high-resolution aerial and satellite data, owing to the similarity of these data to the large datasets in computer vision. Accordingly, this study presents a detailed investigation of state-of-the-art deep learning tools for classification of complex wetland classes using multispectral RapidEye optical imagery. Specifically, we examine the capacity of seven well-known deep convnets, namely DenseNet121, InceptionV3, VGG16, VGG19, Xception, ResNet50, and InceptionResNetV2, for wetland mapping in Canada. In addition, the classification results obtained from deep CNNs are compared with those based on conventional machine learning tools, including Random Forest and Support Vector Machine, to further evaluate the efficiency of the former to classify wetlands. The results illustrate that the full-training of convnets using five spectral bands outperforms the other strategies for all convnets. InceptionResNetV2, ResNet50, and Xception are distinguished as the top three convnets, providing state-of-the-art classification accuracies of 96.17%, 94.81%, and 93.57%, respectively. The classification accuracies obtained using Support Vector Machine (SVM) and Random Forest (RF) are 74.89% and 76.08%, respectively, considerably inferior relative to CNNs. Importantly, InceptionResNetV2 is consistently found to be superior compared to all other convnets, suggesting the integration of Inception and ResNet modules is an efficient architecture for classifying complex remote sensing scenes such as wetlands.

scholarly journals Web-SpikeSegNet: Deep Learning Framework for Recognition and Counting of Spikes from Visual Images of Wheat Plants

2021 ◽  
Author(s):  
Tanuj Misra ◽  
Alka Arora ◽  
Sudeep Marwaha ◽  
Ranjeet Ranjan Jha ◽  
Mrinmoy Ray ◽  
...  
Keyword(s):  
Deep Learning ◽  
Agricultural Research ◽  
Indian Agricultural Research Institute ◽  
Wheat Plant ◽  
Plant Phenotyping ◽  
Visual Images ◽  
Type I ◽  
Spike Detection ◽  
Learning Framework ◽  
Detection And Counting

Abstract Background: Computer vision with deep-learning is emerging as a major approach for non-invasive and non-destructive plant phenotyping. Spikes are the reproductive organs of wheat plants. Detection of spike helps in identifying heading, and counting of the spikes as well as area of the spikes will be useful for determination of the yield of wheat plant. Hence detection and counting of spikes which is considered as the grain bearing organ, has great importance in the phenomics study of large sets of germplasms. Results: In the present study, we developed an online platform “Web-SpikeSegNet” based on a deep-learning framework for spike detection and counting from the wheat plant’s visual images. This platform implements the “SpikeSegNet” approach developed by Misra et al.(2020), which has proved as an effective and robust approach for spike detection and counting. Architecture of the Web-SpikeSegNet consists of 2 layers. First Layer, Client Side Interface Layer, deals with deals with end user’s requests and its corresponding responses management while the second layer, Server Side Application Layer consisting of spike detection and counting module. The backbone of the spike detection module comprises of deep encoder-decoder network with hourglass for spike segmentation. Spike counting module implements the “Analyze Particle” function of imageJ to count the number of spikes. For evaluating the performance of Web-SpikeSegNet, we acquired wheat plant’s visual images using LemnaTec imaging platform installed at Nanaji Deshmukh Plant Phenomics Centre, ICAR-Indian Agricultural Research Institute, New Delhi, India and the satisfactory segmentation performances were obtained as Type I error 0.00159, Type II error 0.0586, Accuracy 99.65%, Precision 99.59% and F1 score 99.65%. Conclusions: In this study, freely available web-based software has been developed based on combined digital image analysis and deep learning techniques. As spike detection and counting in wheat phenotyping are closely related to the yield, Web-SpikeSegNet is a significant step forward in the field of wheat phenotyping and will be very useful to the researchers and students working in the domain.

Sign in / Sign up

Export Citation Format

Share Document