检索结果-内蒙古大学图书馆

EmergencyNet: Efficient Aerial image Classification for Drone-Based Emergency Monitoring Using Atrous Convolutional Feature Fusion

引用

IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING 2020年 13卷 1687-1699页

作者： Kyrkou, Christos Theocharides, Theo Univ Cyprus KIOS Res & Innovat Ctr Excellence CY-1678 Nicosia Cyprus KIOS Res & Innovat Ctr Excellence CY-1678 Nicosia Cyprus Dept Elect & Comp Engn CY-1678 Nicosia Cyprus

deep learning-based algorithms can provide state-of-the-art accuracy for remote sensing technologies such as unmanned aerial vehicles (UAVs)/drones, potentially enhancing their remote sensing capabilities for many emergency response and disaster management applications. In particular, UAVs equipped with camera sensors can operating in remote and difficult to access disaster-stricken areas, analyze the image and alert in the presence of various calamities such as collapsed buildings, flood, or fire in order to faster mitigate their effects on the environment and on human population. However, the integration of deep learning introduces heavy computational requirements, preventing the deployment of such deep neural networks in many scenarios that impose low-latency constraints on inference, in order to make mission-critical decisions in real time. To this end, this article focuses on the efficient aerial image classification from on-board a UAV for emergency response/monitoring applications. Specifically, a dedicated Aerial image Database for Emergency Response applications is introduced and a comparative analysis of existing approaches is performed. Through this analysis a lightweight convolutional neural network architecture is proposed, referred to as EmergencyNet, based on atrous convolutions to process multiresolution features and capable of running efficiently on low-power embedded platforms achieving upto higher performance compared to existing models with minimal memory requirements with less than accuracy drop compared to state-of-the-art models.

关键词： Computer architecture Feature extraction Emergency services Task analysis Remote sensing Convolutional neural networks Monitoring Convolutional neural networks (CNN) deep learning drones emergency monitoring image processing remote sensing unmanned aerial vehicles (UAVs) video processing

来源：评论

学校读者我要写书评

暂无评论

WV-LUT: Wide Vision Lookup Tables for real-time Low-Light image Enhancement

引用

IEEE Transactions on Multimedia 2025年

作者： Li, Canlin Su, Haowen Tan, Xin Zhang, Xiangfei Ma, Lizhuang Zhengzhou University of Light Industry School of Computer Science and Technology Zhengzhou450002 China East China Normal University School of Computer Science and Technology Shanghai200062 China Shanghai Jiao Tong University Department of Computer Science and Engineering Shanghai200240 China

In recent years, the lookup tables (LUTs) with deep learning for image enhancement have achieved remarkable results with extremely high inference efficiency. However, when dealing with severely degraded low-light images, lookup-table-based methods tend to exhibit poor enhancement results due to the lack of contextual and global information. To address the limitations of current lookup-table-based methods in the low-light image enhancement task, we propose the novel Wide Vision Lookup Tables (WV-LUT) by introducing Complementary-Hierarchical 4D-LUTs into 3D-LUT, which allows 3D-LUT to have a wider range of vision. Specifically, the 4D-LUTs are used to expand the receptive field and process local information on a single channel, while a 3D-LUT is used for sRGB channel post-processing. Additionally, we propose a lightweight Global Adjustment Module that further enhances the performance and generalization of WV-LUT by obtaining global adjustment parameters for gamma and color correction matrix to adaptively process images. Experimental results demonstrate that our method outperforms other state-of-the-art methods in low-light image enhancement with the highest average ranking and superior inference efficiency. Furthermore, deployment experiments on mobile devices demonstrate that our WV-LUT achieves superior results and inference efficiency, showcasing promising application prospects for edge devices. The code will be available at https://***/ssw219/WV-LUT. © 1999-2012 IEEE.

关键词： Table lookup

来源：评论

学校读者我要写书评

暂无评论

Application of Multilayer Cascade Fusion Enhanced Network Model (MLCF-Net) to Underwater image processing

Application of Multilayer Cascade Fusion Enhanced Network Mo...

引用

Power Electronics, Computer Applications (ICPECA), IEEE International Conference on

作者： Kexuan Lin Xiran Sun Xi'an Jiaotong University Xi'an China

ISBN: (数字)9798331533694

ISBN: (纸本)9798331533700

In this paper, a multilayer cascade fusion enhancement network model (MLCF-Net) is proposed with the aim of improving the quality of underwater images. Firstly, by classifying the degradation phenomena of underwater images, the study analyzes the causes of bias, blur and low light, and quantifies the characteristics of different degraded images using texture-aware joint color histogram analysis. Based on this, for each degradation case, this paper designs corresponding enhancement strategies: using white balance correction technique to resolve the off-color, Laplace filtering with deep learning super-resolution model to recover blurring details, and using adaptive histogram equalization (CLAHE) to enhance low-light images. Finally, MLCF-Net effectively copes with the coexistence of multiple degradation phenomena in complex scenes through three modules-color correction, texture enhancement, and luminance compensation-combined with a cascade fusion strategy.

关键词： Degradation Histograms Adaptation models Systematics image color analysis Computational modeling Superresolution Nonhomogeneous media real-time systems Power electronics

来源：评论

学校读者我要写书评

暂无评论

Accurate long-term air temperature prediction with Machine learning models and data reduction techniques

引用

APPLIED SOFT COMPUTING 2023年 136卷

作者： Fister, D. Perez-Aracil, J. Pelaez-Rodriguez, C. Del Ser, J. Salcedo-Sanz, S. Univ Alcala Dept Signal Proc & Commun Madrid 28805 Spain Basque Res & Technol Alliance BRTA TECNALIA Derio 48160 Spain Univ Basque Country UPV EHU Bilbao 48013 Spain

In this paper, three customised Artificial Intelligence (AI) frameworks, considering deep learning, Machine learning (ML) algorithms and data reduction techniques, are proposed for a problem of longterm summer air temperature prediction. Specifically, the prediction of the average air temperature in the first and second August fortnights, using input data from previous months, at two different locations (Paris, France) and (Cordoba, Spain), is considered. The target variable, mainly in the first August fortnight, can contain signals of extreme events such as heatwaves, like the heatwave of 2003, which affected France and the Iberian Peninsula. Three different computational frameworks for air temperature prediction are proposed: a Convolutional Neural Network (CNN), with video-to-image translation, several ML approaches including Lasso regression, Decision Trees and Random Forest, and finally a CNN with pre-processing step using Recurrence Plots, which convert time series into images. Using these frameworks, a very good prediction skill has been obtained in both Paris and Cordoba regions, showing that the proposed approaches can be an excellent option for seasonal climate prediction problems. (c) 2023 The Authors. Published by Elsevier B.V.

关键词： deep learning Temperature prediction Recurrence plots Data reduction techniques

来源：评论

学校读者我要写书评

暂无评论

Memristive Quantized Neural Networks: A Novel Approach to Accelerate deep learning On-Chip

引用

IEEE TRANSACTIONS ON CYBERNETICS 2021年第4期51卷 1875-1887页

作者： Zhang, Yang Cui, Menglin Shen, Linlin Zeng, Zhigang Shenzhen Univ Natl Engn Lab Big Data Syst Comp Technol Guangdong Key Lab Intelligent Informat Proc Comp Vis InstSch Comp Sci & Software Engn Shenzhen 518060 Peoples R China Univ Massachusetts Dept Elect & Comp Engn Amherst MA 01003 USA Univ Nottingham Sch Comp Sci Ningbo 315100 Peoples R China Shenzhen Univ Sch Comp Sci & Software Engn Comp Vis Inst Natl Engn Lab Big Data Syst Comp Technol Shenzhen 518060 Peoples R China Shenzhen Univ Guangdong Key Lab Intelligent Informat Proc Shenzhen 518060 Peoples R China Huazhong Univ Sci & Technol Sch Artificial Intelligence & Automat Wuhan 430074 Peoples R China Huazhong Univ Sci & Technol Educ Minist China Key Lab Image Proc & Intelligent Control Wuhan 430074 Peoples R China

Existing deep neural networks (DNNs) are computationally expensive and memory intensive, which hinder their further deployment in novel nanoscale devices and applications with lower memory resources or strict latency requirements. In this paper, a novel approach to accelerate on-chip learning systems using memristive quantized neural networks (M-QNNs) is presented. A real problem of multilevel memristive synaptic weights due to device-to-device (D2D) and cycle-to-cycle (C2C) variations is considered. Different levels of Gaussian noise are added to the memristive model during each adjustment. Another method of using memristors with binary states to build M-QNNs is presented, which suffers from fewer D2D and C2C variations compared with using multilevel memristors. Furthermore, methods of solving the sneak path issues in the memristive crossbar arrays are proposed. The M-QNN approach is evaluated on two image classification datasets, that is, ten-digit number and handwritten images of mixed National Institute of Standards and Technology (MNIST). In addition, input images with different levels of zero-mean Gaussian noise are tested to verify the robustness of the proposed method. Another highlight of the proposed method is that it can significantly reduce computational time and memory during the process of image recognition.

关键词： Memristors Neural networks Acceleration System-on-chip Training Memory management Device-to-device communication Acceleration crossbar array image processing image recognition memristor quantized convolutional neural networks (CNNs)

来源：评论

学校读者我要写书评

暂无评论

Harnessing AI for Agriculture: Plant Pest Detection on Web Application Using deep learning 10

Harnessing AI for Agriculture: Plant Pest Detection on Web A...

引用

10th International Conference on Digital Arts, Media and Technology and 8th ECTI Northern Section Conference on Electrical, Electronics, Computer and Telecommunications Engineering, ECTI DAMT and NCON 2025

作者： Nuankaew, Wongpanya S. Yangwattana, Yotsapron Kamwaree, Sasikan Sararat, Thapanapong Nuankaew, Pratya University of Phayao School of Information and Communicatin Technology Phayao56000 Thailand

ISBN: (纸本)9798331543273

This research develops a mobile web application for pest identification using deep learning to help farmers manage pest outbreaks. The system enables farmers to capture pest images with their mobile cameras and receive alerts and recommendations. The objectives include developing pest classification models, creating web applications, and assessing user satisfaction. YOLOv5 architecture was tested for accuracy, speed, and efficiency. Developed using Flask and Python with SQLite for data management, results showed YOLOv5m6 achieved the highest Recall value of 0.8281 in 4.007 hours, followed by YOLOv5m at 0.8234 in 3.622 hours, and YOLOv5l at 0.8261 in 19.424 hours. Despite similar Recall values, YOLOv5m6 was more efficient. The system demonstrates AI's potential to improve pest management, though challenges in real-time processing and image complexity remain. © 2025 IEEE.

关键词： deep learning

来源：评论

学校读者我要写书评

暂无评论

YoloV4-based defect detection for 3D printing of dental moulds

YoloV4-based defect detection for 3D printing of dental moul...

引用

Chinese Automation Congress (CAC)

作者： Ying Ma Jiang Song Wei Chen Jianxing Li Maoyuan Ji Shusen Kuang Research and Development Center for Industrial Automation Technology of Fujian Province Fujian University of Technology Fuzhou China

Due to the quality defects that occur in the current dental mold 3D printing process affecting the quality of the finished product and the existing algorithms' inability to detect defects in real time and low accuracy, a real-time defect detection scheme for dental mold 3D printing based on YOLOv4 deep learning model is proposed. The 3D printer and the design vision system built to print the dental mold acquire the image dataset of drawing and longitudinal defects, and turn the image dataset into a series of processing into the dataset required for YOLOv4 deep learning for training, prediction and real-time detection. After experimental comparison, YOLOv4 has the highest average detection accuracy of 93.74% and the fastest average detection time of 22 ms among the Faster-RCNN, EfficientDet, YOLOv3 and YOLOv4 deep learning network models. The results show that the method is capable of detecting defects in the 3D printing process in real time, and is advanced and practical for 3D printing defect detection.

关键词： deep learning Training Solid modeling Three-dimensional displays Production Three-dimensional printing real-time systems

来源：评论

学校读者我要写书评

暂无评论

deep learning based Automatic Indoor Plant Management System using Arduino

Deep learning based Automatic Indoor Plant Management System...

引用

Global Conference for Advancement in Technology (GCAT)

作者： Sanjana Prasad K Umme Haniya Syeda Zeenathunnisa Mohammed Tahir Mohammed Sulaiman Dept. of Electronics and Communication Engineering HKBK College of Engineering Bengaluru India Department of Electronics and Communication Engineering HKBK College of Engineering Bengaluru India

Normal plant management systems relies majorly on human monitoring and care. Some of the disadvantages are uneven watering, inadequate nutrition supply to the plants, time consuming, lack in precision of plant health parameters, not having real time visibility, inefficient resource usage of water and energy. In order to overcome these drawbacks, a deep learning-based Plant management system is proposed in this work. By automating the system, there are several advantages. In this project, an Arduino Based Plant Automation System that waters plants automatically based on the amount of soil moisture detected by the sensors is designed. The device provide fertilizer based on detecting plant diseases through image processing software. Some advantages of the proposed system include Accurate Monitoring of plants health parameters. Automated watering, fertilizing, providing real time alerts and notifications, improved plant health and growth, energy and resource efficiency, remote access and control etc. Some areas where this system can be deployed are Smart homes, Indoor gardening, Horticulture research, Sustainable Agriculture and Vertical Farming.

关键词：

来源：评论

学校读者我要写书评

暂无评论

X-Ray image-Based real-time COVID-19 Diagnosis Using deep Neural Networks (CXR-DNNs)

引用

JOURNAL OF IMAGING 2024年第12期10卷 328-328页

作者： Khan, Ali Yousuf Luque-Nieto, Miguel-Angel Saleem, Muhammad Imran Nava-Baro, Enrique Univ Malaga Telecommun Engn Sch Malaga 29010 Spain Univ Malaga Inst Ocean Engn Res Malaga 29010 Spain Sir Syed Univ Engn & Technol Dept Software Engn Karachi 75300 Pakistan

On 11 February 2020, the prevalent outbreak of COVID-19, a coronavirus illness, was declared a global pandemic. Since then, nearly seven million people have died and over 765 million confirmed cases of COVID-19 have been reported. The goal of this study is to develop a diagnostic tool for detecting COVID-19 infections more efficiently. Currently, the most widely used method is Reverse Transcription Polymerase Chain Reaction (RT-PCR), a clinical technique for infection identification. However, RT-PCR is expensive, has limited sensitivity, and requires specialized medical expertise. One of the major challenges in the rapid diagnosis of COVID-19 is the need for reliable imaging, particularly X-ray imaging. This work takes advantage of artificial intelligence (AI) techniques to enhance diagnostic accuracy by automating the detection of COVID-19 infections from chest X-ray (CXR) images. We obtained and analyzed CXR images from the Kaggle public database (4035 images in total), including cases of COVID-19, viral pneumonia, pulmonary opacity, and healthy controls. By integrating advanced techniques with transfer learning from pre-trained convolutional neural networks (CNNs), specifically InceptionV3, ResNet50, and Xception, we achieved an accuracy of 95%, significantly higher than the 85.5% achieved with ResNet50 alone. Additionally, our proposed method, CXR-DNNs, can accurately distinguish between three different types of chest X-ray images for the first time. This computer-assisted diagnostic tool has the potential to significantly enhance the speed and accuracy of COVID-19 diagnoses.

关键词： COVID chest X-ray images image classification deep learning vision transformer lung infection

来源：评论

学校读者我要写书评

暂无评论

A Survey on Semantic Segmentation of Aerial images using deep learning Techniques 3

A Survey on Semantic Segmentation of Aerial Images using Dee...

引用

3rd IEEE International Virtual Conference on Innovations in Power and Advanced Computing Technologies, i-PACT 2021

作者： Anilkumar, P. Venugopal, P. School of Electronics Engineering Vellore Institute of Technology Tamil Nadu Vellore632014 India

ISBN: (纸本)9781665426916

Semantic segmentation of remote sensing images has been utilized in an assortment of uses and has been a focal point of examination for a long time. With the achievement of profound learning strategies in PC vision, analysts have endeavored to move their better presentation than the field of distant detecting picture examination. In the field of computer vision, semantic segmentation is an interesting task. It is the method of assigning every pixel to a specific label. Aerial images have become more common in all fields of science as a result of technological advancements in recent years. Tables, maps, graphs, and other forms of crucial data for somewhat area could be attained, but these are insufficient to present a real-time analysis. As a result, an aerial picture seals in the gaps. To improve the quality of the images, they must go through several processing steps, Segmentation is one example of such processing. The foremost objective of image segmentation is to group the pixels of areas analogous to distinct planes, items, or usual portions of things in order to streamline or alter the illustration of an image into somewhat additional significant and easier to analyse. The availability of very high-resolution remote sensing images, combined with advances in deep learning methods, has moved the standard of image taxonomy left from pixel-based and object based approaches and toward deep learning-based semantic segmentation. In this paper, we presented a review of some cutting-edge Convolutional Neural Network (CNN) architectures such as AlexNet, GoogleNet, VGGNet, and ResNet that serve as the foundation for Semantic Segmentation, as well as a revision of numerous segmentation procedures applied to aerial images. © 2021 IEEE.

关键词： Semantic Segmentation

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：