检索结果-内蒙古大学图书馆

A non-intrusive approach for fault detection and diagnosis of water distribution systems based on image sensors, audio sensors and an inspection robot

引用

ENERGY AND BUILDINGS 2021年 243卷 110967-110967页

作者： He, Ruikai Xu, Peng Chen, Zhibo Luo, Wei Su, Zhineng Mao, Jiong Tongji Univ Dept Mech & Energy Engn Room A4344800 Caoan Rd Shanghai 201804 Peoples R China Shanghai Hongqiao Int Airport Shanghai 201804 Peoples R China Shanghai Dynawin Facil Management Co Ltd Shanghai 201804 Peoples R China

Fault diagnosis is important to maintain the normal operation of air-conditioning systems, reduce the energy consumption in buildings, and increase the service life of air-conditioning system equipment. We present a novel approach for fault detection and diagnosis system that relies on image and audio sensors and relevant algorithms. This paper proposes a fault diagnosis algorithm based on a robot that can automatically capture audio and image signals from microphone arrays and cameras during inspection in a chiller room. It includes audio-and image-based fault diagnosis algorithms. The validity of the algorithm combined with sensors is verified using data from actual equipment in a chiller room. The audio-based algorithm, which can monitor the abnormal sound of pumps to detect faults, utilizes Fourier transform, a finite impulse response digital filter, and an autoregressive integrated moving average model. We analyze the frequency domain of the pump signal and set the appropriate threshold to monitor abnormal signals based on the fitted model. Meanwhile, the image-based algorithms are divided into three sections to achieve three functions: 1) an AlexNet convolutional neural network is modified to classify the images of the chiller room equipment obtained by the visible light camera;2) image morphology methods and trigonometric functions are used to read the dials' indicators acquired by the visible light camera;and 3) optical character recognition is used to obtain the highest temperature value in the infrared image of the pump captured by the infrared camera, which helps maintenance staff verify the operation of the pump and detect faults as soon as possible. These diagnostic algorithms are non-intrusive, low cost, and easy to deploy. Combined with real-time data collection from the sensors on the robot, the algorithms can effectively improve the intelligence of the equipment room and allocate human resources more reasonably. (c) 2021 Elsevier B.v. All rights res

关键词： Fault diagnosis Audio signal processing image processing Inspection robot

来源：评论

学校读者我要写书评

暂无评论

v2X Communications for Platooning: Impact of Sensor Inaccuracy

V2X Communications for Platooning: Impact of Sensor Inaccura...

引用

11th International image processing and Communications Conference (IP and C)

作者： Sybis, Michal Sroka, Pawel Kliks, Adrian Kryszkiewicz, Pawel Poznan Univ Tech Pl M Sklodowskiej Curie 5 PL-60965 Poznan Poland

ISBN: (纸本)9783030312541;9783030312534

Platooning is one of the prospective applications, where autonomous driving and vehicle-to-vehicle and vehicle-to-Infrastructure communications play a vital role. various test proved that reduction of inter-car distance withing platoon of trucks results in significant fuel savings. Keeping of such short distance as well as awareness of the platooning cars about the vicinity assumes the presence of highly reliable sensors. In this paper, we briefly discuss the impact of sensor inaccuracy on the overall behaviour of the train-of-trucks driving in the autonomous fashion, supported by the presence of dedicated wireless system utilizing context information stored in databases and maps.

关键词： v2X and 5G Sensor accuracy Context information database

来源：评论

学校读者我要写书评

暂无评论

A Personalized Recommendation System by Incorprating Graph Neural Network with Attention Mechanism

Research Square

引用

Research Square 2022年

作者： Wilson, Mark Korshin, James Pollard, Bernd Leckie, Krik University of Leeds United Kingdom

As an important link in the information service of Internet products, the recommendation system has become an important way for users to get information from the huge amount of Internet data. Among existing studies, collaborative filtering has become a widely used method in recommender systems due to its good performance, but it uses a shallow model that cannot learn the deep non-linear features of users and items. In addition, content-based recommendation methods make recommendations by making full use of user registration information and item profiles, but this method also requires effective feature extraction, relying on feature engineering i.e., by manually extracting or design features, which makes the effectiveness and scalability of the method very limited and limits the performance of the recommendation algorithm. In recent years, deep learning has made a great impact in natural language processing, speech recognition, and image processing. This has led to a new breakthrough in the research of recommendation systems. Graph neural network algorithms extend traditional deep learning methods, such as convolution, to the domain of graph data, and combine the idea of data propagation to form deep learning algorithms on graphs, which have achieved good results in social networks, recommendation systems, knowledge graphs, and other fields. In this paper, we focus on the shallow output of the intermediate layer output. A multi-headed attention mechanism is used to fuse the multi-layer output information, so that the shallow structural features provided by the intermediate layer can be better involved in the scoring prediction task, which makes the application of graph neural networks in recommendation systems further improved. © 2022, CC BY.

关键词： Recommender systems

来源：评论

学校读者我要写书评

暂无评论

Using Machine Learning Techniques to Increase the Effectiveness of Cybersecurity

Using Machine Learning Techniques to Increase the Effectiven...

引用

2021 Cybersecurity Providing in Information and Telecommunication systems II, CPITS-II-2 2021

作者： Buhas, vasyl Ponomarenko, Ihor Bugas, vаlеriy Ramskyi, Andrii Sokolov, volodymyr Kyiv National University of Technologies and Design 2 Nemyrovycha-Danchenka str. Kyiv01011 Ukraine Borys Grinchenko Kyiv University 18/2 Bulvarno-Kudriavska str. Kyiv04053 Ukraine

In today's world, a great number of organizations generate and accumulate large amounts of information, which is of great value to owners, and is also considered by attackers as a valuable resource for enrichment. Any data storage system has vulnerabilities that will be exploited during cyberattacks. The inability to build a system secure enough against unauthorized access to data, forces companies to respond on an ongoing basis to evolving technologies of misappropriation of information by developing more effective methods of identifying and combating cyberattacks. This article examines the features of the use of machine learning methods to identify illegal access by third parties to the information of individuals and legal entities with economic and reputational damage. The study considers methods of processing various types of data (numerical values, textual information, video and audio content, images) that can be used to build an effective cybersecurity system. Obtaining a high level of identification of unauthorized access to data and combating their theft is possible through the implementation of modern machine learning approaches, which are constantly improving by creating innovative data processing algorithms and the use of powerful cloud computing services, acting as an element to counter rapidly evolving technologies. © 2022 Copyright for this paper by its authors.

关键词： image recognition

来源：评论

学校读者我要写书评

暂无评论

Machine Learning-Based System for Detecting and Tracking Driver Drowsiness

Machine Learning-Based System for Detecting and Tracking Dri...

引用

International Conference on Smart Structures and systems (ICSSS)

作者： S.v. Priya Archana. M Bhuvaneshwari S Niranjana Bharathee R Electronics and Communication Engineering Loyola Institute of Technology

Drivers who aren't paying attention or staying attentive are to blame for many accidents occurring on the world's roads nowadays. The phrase for this condition is "driver drowsiness." This results in a wide variety of undesirable circumstances that endanger human lives and harm them. The key goals of this study are to identify driver sleepiness and to design an appropriate reaction to such findings. There are several ways, many based on the vehicle's movement or the driver's actions. One of the approaches is the physiological method, which assists in diverting the driver's attention away from exhaustion and making him alert. Only some of the solutions entail the use of expensive sensors or a considerable quantity of data. Consequently, this research aims to develop a system that can identify sleepiness in real-time while maintaining an acceptable degree of accuracy. In this method, a camera is used to capture the driver's facial expressions so that they may be examined afterward. A variety of image processing methods are employed to assess each frame for all motion. The landmark locations on the face are used to calculate the Eye Aspect Ratio (EAR), Nose Length Ratio (NLR) & Mouth Opening Ratio(MOR). The derived values are compared to the system's threshold values, and the detection is based on the difference in weight. While this is happening, machine learning algorithms are also being applied offline. Based on the categorization, the system achieved a sensitivity of 98.45% and a specificity of 100% using a Random Forest Algorithm (RFA). This model system is appropriate for use with all sorts of motor vehicles

关键词：

来源：评论

学校读者我要写书评

暂无评论

Priority Scheduling Using Recurrent Quadrant Search for Handling Priority and ‘Pop-Up’ Targets in Aerial videos

Priority Scheduling Using Recurrent Quadrant Search for Hand...

引用

Applied imagery Pattern Recognition Workshop (AIPR)

作者： Chandrakanth. v Shivpal Singh Sumohana S Channappayya K Palaniappan Defence Research and Development Laboratory Hyderabad Indian Institute of Technology Hyderabad University of Missouri Columbia

Target detection and tracking is a well established application in the radar (RF) domain. However, the same operating principles may not directly extend to image-based tracking systems. In the optical domain, the problem becomes challenging due to the lack of an active transmit signal (as available in radar processing), which actively maps the environment for target detection and track initiation. Passive imaging systems often require a separate assisted intervention in the optical processing channel for real-time target detection and track initialization. This limitation has been successfully addressed with the development of automated tracking algorithms using deep learning architectures like Convolutional Neural Networks (CNNs). CNNs are extensively employed in generic object classification and localization applications for single and multi-target scenarios. In multi-target scenarios, the deep architecture framework detects all the objects in the frame without any priority designation (eg. YOLO, vGG19 etc.). However, in multi-target scenarios where, handling selective targets is critical (eg. tracking an ambulance in traffic) or prioritized tracking is needed of a ‘Pop-Up’ target, that is an object entering the field-of-view (FOv)) (i.e. in autonomous driving systems, e-commerce deliveries etc.), the generic networks are not sufficient. In this work, we aim at addressing this specific problem of priority/ selective target tracking in aerial videos. We propose a novel CNN termed ‘PriorityNet’ with embedded priority information and a recurrent quadrant search algorithm for prioritized tracking and false alarm rejection in multi-target scenarios. The CNN and the associated algorithm is successfully deployed on a portable NvIDIA Jetson-Nano board (a small form factor GPU) integrated on-board a custom built UAv. The complete system is tested on multiple open-source and custom multi-modal data sets (RGB and Thermal). The results obtained demonstrate the efficacy of the

关键词：

来源：评论

学校读者我要写书评

暂无评论

Deep Learning and IoT Based Robotics to Monitor the Traffic 2nd

Deep Learning and IoT Based Robotics to Monitor the Traffic

引用

2nd International Conference on Cyber Warfare, Security and Space Computing, SpacSec 2024

作者： Priya, v. vishwa Pande, Soumitra S. Ilyas, Md Jayasudha, R. Ramesh, Janjhyam venkata Naga Suganthi, D. Department of Computer Science Vels Institute of Science Technology and Advanced Studies Chennai India Ericsson Global India Private Limited Bangalore India Prestige Institute of Engineering Management and Research Indore India Dr. N. G. P. Institute of Technology Coimbatore India Koneru Lakshmaiah Education Foundation Andhra Pradesh Guntur Vaddeswaram India Saveetha College of Liberal Arts and Sciences SIMATS Thandalam Chennai India

ISBN: (纸本)9783031734939

In densely populated areas, managing traffic laws has become more difficult due to the rising number of cars on the road. One of the most critical components of effective traffic management and improved mobility is real-time traffic monitoring systems. Consequently, precise and dependable real-time traffic information has always been essential for cars and drivers. Many other approaches have been suggested as potential answers to the issues plaguing traffic conditions recently. A different approach is vCC or vehicle cloud computing. In addition, an advanced model called an Iot-aided robotic (IoRT), which uses cameras and Internet of Things (IoT) detector nodes to gather actual traffic information, has been created. Two deep learning methods, one based on improved LeNet-5 for real-time signs for traffic identification and the other on the Inception-v3 model for detecting and identifying traffic lights, are the critical contributions provided by this research effort. Additionally, to decrease road accidents, the ideal distance between barriers and ultrasonic sensors was determined by analyzing the timing and speed of ultrasonic waves. Data acquired by cameras and sensors is analyzed using a variety of image processing algorithms before being uploaded to the cloud and made accessible to commuters and drivers via a mobile app. According to the test findings, the suggested models are far more accurate. The expanded GTSRB (EGTSRB) dataset achieved 98.89% accuracy while the German Traffic Sign Recognition Benchmark (GTSRB) dataset achieved 98.23% accuracy using the modified LeNet-5. The second model achieved a precision of 99.7 percent using data from the Laboratory for the Intelligent and Safe Automobiles (LISA). The results of this study beat previous research on comparable traffic monitoring systems by 2.13% for traffic light recognition and detection and by 4.89% for traffic sign identification. © The Author(s), under exclusive license to Springer Nature Switzerland AG

关键词： Traffic signs

来源：评论

学校读者我要写书评

暂无评论

Methods and algorithms for processing a series of cloud images when observing from the Earth's surface 26

Methods and algorithms for processing a series of cloud imag...

引用

26th International Symposium on Atmospheric and Ocean Optics, Atmospheric Physics

作者： Galileiskii, v. P. Elizarov, A., I Kokarev, D., v Morozov, A. M. VE Zuev Inst Atmospher Opt SB RAS Tomsk Russia

ISBN: (纸本)9781510639423

The article discusses the methods and algorithms used in the development of software and hardware systems for registration and analysis of cloud images, when observed from the surface of the Earth. The methods of pre-processing images are considered and the possibility of determining the direction and apparent angular velocity of cloud movement based on a series of images for the purpose of making a short-term forecast is evaluated.

关键词： image processing cloudiness allsky camera camera calibration optical flow

来源：评论

学校读者我要写书评

暂无评论

Adaptive distributed video surveillance system

Adaptive distributed video surveillance system

引用

2020 International Conference on Technology and Entrepreneurship - virtual, ICTE-v 2020

作者： Khelvas, Alexander Demyanova, Darya Gilya-Zetinov, Alexander Konyagin, Egor Khafizov, Roman Pashkov, Ruslan MIPT Center of Open Systems Dolgoprudnii Russia Skoltech Center of Open Systems Moscow Russia

ISBN: (纸本)9781728193748

video surveillance systems are a major multipurpose video data source for modern emergency management solutions. As their complexity and image quality grows, the requirements for data exchange channels, storage and processing units are becoming much more ***, a wide range of neural networks becomes a universal instrument for image analysis. Such algorithms require deployment of powerful HPC-class server *** this article, we propose a new adaptive approach to decentralized environment for video data storage and event-driven video processing for modern emergency management *** proposed solutions have been tested on different arrays of video data collected from various sources. These sources include city video surveillance systems, supermarket video surveillance systems, *** each of the listed examples, a discrete dynamic model has been developed based on video processing and event *** research resulted in a new data collection and analysis approach based on distributed video processing, reconstruction of 3D scenes and tactical situations detection. © 2020 IEEE.

关键词： Electronic data interchange

来源：评论

学校读者我要写书评

暂无评论

Fused Feature vector based DenseNet201 with Deep Convolutional Generative Adversarial Network for Forest Fire Prediction

Fused Feature Vector based DenseNet201 with Deep Convolution...

引用

Software, systems and Information Technology (SSITCON), International Conference on

作者： Mohammed Hussein Fallah Somireddy Spandana M Parameswari G. Saritha Abhijeet Das Department of Computers Techniques Engineering College of Technical Engineering The Islamic University Najaf Iraq School of Computer Science and Engineering Malla Reddy Engineering College for Women (UGC-Autonomous) Maisammaguda Hyderabad India Department of Information Technology Sona College of Technology Salem India Department of Electronics and Communication Engineering Sri Sai Ram Institute of Technology Chennai India Department of Civil Engineering C.V. Raman Global University (CGU) Bhubaneswar Odisha India

ISBN: (数字)9798350352931

ISBN: (纸本)9798350352948

The rise in temperature on the earth is one of the main sources for the cause of fire in the forest area. The dry vegetation of the forest leads to sudden fire in the forest. various technologies have been introduced to capture the aerial views of the forest to monitor the ongoing actions. The different existing methods failed to predict the occurrence of the fire in the forest area due to lack of feature dimensionalities handling and noises generated during the training process that reduced the performance of appropriate forest fire prediction. To overcome the existing problems the Fused Feature vector Based DenseNet201 (FFv-DenseNet201) is proposed with the Deep Convolutional Generative Adversarial Network (DCGAN). The min-max normalization was used in the pre-processing stage to rescale the image pixel ranges. The FFv-DenseNet201 is used to extract the significant features from the images. The DCGAN algorithm was used to predict the fire and non-fire forest images. The developed FFv-DenseNet 201 and DCGAN has obtained better results of $\mathbf{9 8. 3 4 \%}$ accuracy, $\mathbf{9 9. 0 1 \%}$ of precision, $\mathbf{9 8. 2 9 \%}$ recall and $\mathbf{9 8. 6 4 \%}$ f1-score respectively for the FLAME dataset compared to existing YOLOv5 model.

关键词： YOLO Accuracy Fires vegetation mapping Forestry Predictive models Feature extraction Prediction algorithms Generative adversarial networks vectors

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：