检索结果-内蒙古大学图书馆

SSRN 2024年

作者： Mondal, Shyamal Jampani, Kashyap Raj, Akshay R. Sethi, Abhijit Chowdhury, Dibakar Roy Dept. of Applied Physics Defence Institute of Advanced Technology Maharashtra Pune411025 India Dept. of Physics Mahindra Ecole Centrale University Hyderabad500043 India

Terahertz images typically suffer from poor image quality and do not allow traditional machine learning methods to detect and identify objects. In this work, an image segmentation algorithm called W-Net is deployed for the first time, to segment concealed objects from the poor-quality Terahertz image samples. This involves developing an image enhancement and segmentation system for terahertz images. The proposed system comprises of two stages:in the first stage, the resolution of low-resolution THz images has been enhanced by training a U-Net deep learning model. In the latter stage, another U-Net model is trained to segment the enhanced images using corresponding masks to identify regions of interest. The proposed system is evaluated using a dataset of low-resolution terahertzimages with its respective high-resolution images and masks. The system performed well with an accuracy, precision, receiver operating characteristic and F1 scores of 99.1%, 0.9143, 0.9977 and 0.9146, respectively for enhancement. Moreover, values of the same evaluation metrics deduced as 99.9%, 0.9851, 0.9999 and 0.9923, respectively for segmentation. Therefore, the demonstrated scheme is expected to identify and classify objects hidden in real-time terahertz images and improve the quality of the same to enable accurate segmentation for applications like security screening, biomedical applications, quality and health monitoring in food industry, etc. © 2024, The Authors. All rights reserved.

关键词： image enhancement

来源：评论

学校读者我要写书评

暂无评论

deep learning-based image processing framework for efficient surface litter detection in Computer Vision applications

引用

JOURNAL OF RADIATION RESEARCH AND APPLIED SCIENCES 2025年第2期18卷

作者： Shad, Ifza Zhang, Zuping Asim, Muhammad Al-Habib, Mohammed Chelloug, Samia Allaoua Abd El-Latif, Ahmed Cent South Univ Sch Comp Sci & Engn Changsha 410083 Peoples R China Prince Sultan Univ Coll Comp & Informat Sci EIAS Data Sci Lab Riyadh 11586 Saudi Arabia Prince Sultan Univ Ctr Excellence Quantum & Intelligent Comp Riyadh 11586 Saudi Arabia Princess Nourah bint Abdulrahman Univ Coll Comp & Informat Sci Dept Informat Technol POB 84428 Riyadh 11671 Saudi Arabia Menoufia Univ Fac Sci Dept Math & Comp Sci Shibin Al Kawm 32511 Egypt

Computer vision plays a vital role in automating environmental analysis by enabling real-time object detection and classification in diverse conditions. Litter pollution poses significant health and environmental risks due to inefficient disposal, and manual oversight is labor intensive. Effective litter detection is crucial for large-scale environmental monitoring. However, existing models face challenges such as the complexity of detecting shadowy objects in varying lighting conditions (e.g., during rain or under sun rays), difficulty in recognizing small objects, low accuracy, and poor real-time performance. Existing two-stage detectors, such as Faster R-CNN, also struggle with these issues. This paper introduces an automated deep learning-based image processing approach for accurate litter detection across different locations, using an enhanced version of YOLOv9s called LD-YOLOv9s. Key improvements in this novel approach include replacing convolutional layers with DynConvLayer in the backbone, integrating an SDConv-ADown module to substitute down-sampling layers in the neck, and using MPD-IoU instead of CIoU. These modifications reduce the chances of overlooking small objects, such as caps or lids, which had the least class meaning in the dataset, achieving a mAP of 78.3% with an inference time of 6.7ms. A significant contribution of this work is the LD-2024 dataset, curated from indoor and outdoor environments with manually annotated images. Performance comparisons were made with several YOLO versions (YOLOv3, YOLOv5, YOLOv6, YOLOv7, YOLOv8, YOLOv9 variants) and traditional object detectors (Faster R-CNN, Center Net, Retina Net, Cascade R-CNN). Ablation studies validate the effectiveness of LD-YOLOv9s, which outperforms conventional methods, achieving a 6.3% improvement in mean average precision (mAP) over YOLOv9s on the LD-2024 dataset.

关键词： deep learning Computer vision image processing Litter Detection YOLO DynConvLayer SDConv MPD-IoU LD-YOLOv9s Object Detection

来源：评论

学校读者我要写书评

暂无评论

Using deep learning Approaches for the Automatic Detection of COVID-19 and Assessing Disease Severity through Chest CXR and CT Scan image processing 3

Using Deep Learning Approaches for the Automatic Detection o...

引用

3rd IEEE International Maghreb Meeting of the Conference on Sciences and Techniques of Automatic Control and Computer Engineering, MI-STA 2023

作者： Abubaker Mohamed, Adel Abuzaid, Kenz Universty of Benghazi Benghazi Libya Universty of Benghazi Electrical and Electronics Engineering Department Benghazi Libya Universty of Benghazi Information Technology College Benghazi Libya

ISBN: (纸本)9798350319897

The novel coronavirus has been rapidly spreading globally since its first appearance in China at the end of 2019 and has become a worldwide pandemic. Early detection and isolation of infected cases are major factor in decreasing the spread of the virus. The most popular and effective method for detecting the disease is RT-PCR, but it is expensive and time-consuming. Radiography is a diversty option for detecting the infected cases and determining its severity. However, this process requires experienced radiologists to manually interpret the images. This paper aims to evaluate the effectiveness of using DL models in the automatic classification of the COVID-19 and quantitative measurement of the severity of the spread of the virus in the lungs through CXR and CT scan image processing. Two models have been introduced, one specifically designed for processing CXR images using CNN and DL (Xcov_model), and another based on the structure of the first model but intended for processing CT scan images (CTcov_model). Both models were supported by the Grad-Cam algorithm to create a heat map indicating where the disease is expected to spread. The dataset consisted of 9000 images, which were evenly divided into three classes for both CXR and CT scans. 80% of these images were used for training and 20% for testing, in addition to the implementation of DA technology. The models were implemented and evaluated using Python on the Google Collaboratory platform, with the Xcov_model achieving an F1-Score of 98% for both the COVID-19 class and the normal class, and the CTcov_model having a test accuracy of about 98%. © 2023 IEEE.

关键词： COVID-19

来源：评论

学校读者我要写书评

暂无评论

TPU Based deep learning image Enhancement for real-time Point-of-Care Ultrasound

IEEE TRANSACTIONS ON COMPUTATIONAL IMAGING

引用

IEEE TRANSACTIONS ON COMPUTATIONAL IMAGING 2024年 10卷 461-468页

作者： Huang, Ouwen Palmeri, Mark L. Duke Univ Dept Biomed Engn Durham NC 27708 USA

Handheld ultrasound devices are becoming more prevalent in point-of-care ultrasound workflows. However, these devices are computationally constrained which challenges the advances in deep learning methodology for real-time use on mobile Point-of-care ultrasound (POCUS) devices. In this work, we explore the feasibility of running MimickNet, a deep learning clinical post-processing model, on Tensor processing Units (TPUs), hardware designed for deep learning operations capable of running on only 2 watts of power at 1.8 V with a form factor of 10 mm x 15 mm. We show that real-time deep learning based post-processing is feasible at 20 - 120 FPS for 1472x160 to 224x224 axial sample x B-mode scan line configurations. We refer to the TPU based model as MimickNet Mobile. MimickNet Mobile achieves outputs nearly identical to the original MimickNet with a structural similarity index measurement (SSIM) of 0.98 +/- 0.001 and a mean squared error (MSE) of 0.0001 +/- 0.0 over our test set of 588 frames consisting of 168 phantom frames and 420 prospectively acquired human liver frames. We investigate the latency of other common mobile architectures such as separable convolution. Finally, we investigate the distribution of model parameter error when quantizing MimickNet float32 weights to MimickNet Mobile int8 weights. This work demonstrates that real-time POCUS deep learning image enhancement is feasible using TPUs. Future ultrasound device manufacturers can consider incorporating a TPU for the added flexibility of supporting several deep learning architectures without compromising on power management and form factor.

关键词： Clinical ultrasound deep learning image enhancement MimickNet real-time processing TPU

来源：评论

学校读者我要写书评

暂无评论

Integrating image processing Techniques with deep learning Network Models to Detect and Identify Defects on the Surface of Tomato

Integrating Image Processing Techniques with Deep Learning N...

引用

Electrical, Computer and Energy Technologies (ICECET), International Conference on

作者： Dinh Do van Sao Do University Hai Duong Viet Nam

Before export, fruit should be classified to improve quality, meet customer requirements and increase product value. This article proposes a method to identify defects on the surface of tomato skin using image processing techniques combined with deep learning models. The identification method includes the following main steps: (i) data collection (image of tomato: green, ripe, diseased, scratched), (ii) image labeling, (iii) data file division, (iv) model training, (v) selection and using models. The results of using Faster R-CNN model combining Resnet-10l and testing on YOLOv5 to identify and classify tomatoes that met and failed export for high accuracy (95.3 %) and met get real time.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Application of entertainment interactive robots based on deep learning in referee assistance mode in sports competitions

引用

ENTERTAINMENT COMPUTING 2025年 52卷

作者： Li, Xiaolong Shandong Univ Polit Sci & Law Student Affairs Off Jinan 250014 Shandong Peoples R China

Interactive robots are intelligent auxiliary tools that can monitor sports events in real-time and provide entertainment information services. This article studies the application of entertainment interactive robots based on deep learning in the referee assistance mode of sports competitions. The system first uses a camera to capture real-time images of volleyball matches, then preprocesses the images using image processing algorithms, and uses deep learning algorithms to recognize and track the balls and players in the images. By training the model, the system can accurately determine key information such as player actions. deep learning technology is used to train interactive entertainment robots to identify and analyze key decision events in games. Through image recognition, action analysis and rule matching, the robot monitors the game process in real time and determines the referee's possible errors in the decision. The system generates the penalty results based on the penalty rules and competition rules, and displays them to the referee and audience through display screens or sound prompts. After experimental verification, the volleyball match referee judgment assistance system based on image processing and deep learning has performed excellently in terms of accuracy and speed. Compared to manual referees, the system can identify and track the ball and players more quickly, reduce the possibility of misjudgments, and improve the fairness of the game.

关键词： deep learning Interactive robot for entertainment Sports competitions Referee auxiliary mode

来源：评论

学校读者我要写书评

暂无评论

BLFM-Net:An Efficient Regional Feature Matching Method for Bronchoscopic Surgery Based on deep learning Object Detection

引用

Computers, Materials & Continua 2025年第6期83卷 4193-4213页

作者： He Su Jianwei Gao Kang Kong School of Mechanical Engineering Tianjin University135 Yaguan RoadJinnan DistrictTianjin300350China

Accurate and robust navigation in complex surgical environments is crucial for bronchoscopic *** study purposes a bronchoscopic lumen feature matching network(BLFM-Net)based on deep learning to address the challenges of image noise,anatomical complexity,and the stringent real-time *** BLFM-Net enhances bronchoscopic image processing by integrating several functional *** FFA-Net preprocessing module mitigates image fogging and improves visual clarity for subsequent *** feature extraction module derives multi-dimensional features,such as centroids,area,and shape descriptors,from dehazed *** Faster RCNN Object detection module detects bronchial regions of interest and generates bounding boxes to localize key *** feature matching module accelerates the process by combining detection boxes,extracted features,and a KD-Tree(K-Dimensional Tree)-based algorithm,ensuring efficient and accurate regional feature *** BLFM-Net was evaluated on 5212 bronchoscopic images,demonstrating superior performance compared to traditional and other deep learning-based image matching *** achieved real-time matching with an average frame time of 6 ms,with a matching accuracy of over 96%.The method remained robust under challenging conditions including frame dropping(0,5,10,20),shadowed regions,and variable lighting,maintaining accuracy of above 94%even with the frame dropping of *** study presents BLFM-Net,a deep learning-based matching network designed to enhance and match bronchial features in bronchoscopic *** BLFM-Net shows improved accuracy,real-time performance,and reliability,making a valuable tool for bronchoscopic surgeries.

关键词： Bronchial region feature matching bronchoscopic tracking real-time processing bronchial texture features bronchial texture features deep learning medical image dehazing

来源：评论

学校读者我要写书评

暂无评论

image and Video Captioning Using deep learning and Natural Language processing

Image and Video Captioning Using Deep Learning and Natural L...

引用

International Conference on Computing Communication Control and Automation (ICCUBEA)

作者： Manoj Naidu Athrva Kulkarni Sahil Kadam Siddhesh Joshi Nilesh P. Sable Anuradha Yenkikar Department of CSE – Artificial Intelligence Vishwakarma Institute of Information Technology Pune India

ISBN: (数字)9798350391770

ISBN: (纸本)9798350391787

deep learning models have been a huge success in image recognition which hence can be used for the purpose of text generation. In the field of imaging science, captioning images and videos is regarded as an intellectually difficult job. Visual Geometry Group (VGG); is a standard deep Convolutional Neural Network (CNN) architecture with multiple layers, specifically focusing on the integration of CNN for image feature extraction. Exploring this underlying method, the use of another model is essential for caption generation. Here the Recurrent Neural Network (RNN) comes in use for caption generation from the extracted features. Models named Long Short-Term Memory (LSTM) based on RNN and Bidirectional encoder representation transformer (BERT) based on Transformers have been prominent in ensuring accurate results. The Flicker8k dataset is used which provides a variety of information useful for model training. By testing validation data along with evaluation metrics, we analyze the effectiveness of different models to create consistent and descriptive headlines. Extending our inquiry to encompass title generation using transformer models, while also exploring learning techniques for real-time title generation and delivery using the Open-CV library available in Python to get the output from the camera and display it on screen. The result shows that the LSTM is the best model for captioning, with an accuracy of 65.07% at the epochs of 300 and the BERT model has an accuracy of 31% at the epochs of 2. The findings of this study not only contribute to advancing subtitle enhancement methodologies but also broaden the potential applications of deep learning techniques in this domain.

关键词： deep learning Recurrent neural networks Accuracy Computational modeling Bidirectional control Transformers Feature extraction Encoding Data models Long short term memory

来源：评论

学校读者我要写书评

暂无评论

Distributed deep learning for Medical image processing in Cloud Environments

Distributed Deep Learning for Medical Image Processing in Cl...

引用

Artificial Intelligence for Innovations in Healthcare Industries (ICAIIHI), International Conference on

作者： Neeraj Varshney Parul Madan Anurag Shrivastava C Praveen Kumar A Kakoli Rao Amit Srivastava Department of Computer Engineering and Applications GLA University Mathura Department of Computer Science & Engineering Graphic Era Deemed to be University Dehradun Uttarakhand Saveetha School of Engineering Saveetha Institute of Medical and Technical Sciences Chennai Tamilnadu Department of Computer Science and Engineering Institute of Aeronautical Engineering Hyderabad Telangana Lloyd Institute of Engineering and Technology Greater Noida Lloyd Law College Greater Noida

The incorporation of distributed deep learning for medical image processing in cloud settings is the subject of this study. The findings demonstrate the high viability and significant performance advantages realized by cloud-based distributed systems, notably significant processing time savings, outstanding diagnostic accuracy, as well as improved scalability. The consequences for security and privacy have been discussed, with a focus on effective safeguards for private medical information. There is a void in the literature about resource and cost-effectiveness optimization tactics used in cloud-based systems. Future research must concentrate on resource optimization tactics for economic sustainability, study developing security risks and privacy techniques, and incorporate real-world implementations in order to improve this topic. This study informs the use of distributed deep learning in cloud-based medical image processing as well as adds to the body of knowledge in healthcare technology.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Automatic ventricular nuclear magnetic resonance image processing with deep learning

引用

MULtimeDIA TOOLS AND APPLICATIONS 2021年第26-27期80卷 34103-34119页

作者： Yong, Binbin Wang, Chen Shen, Jun Li, Fucun Yin, Hang Zhou, Rui Lanzhou Univ Sch Informat Sci & Engn Lanzhou Gansu Peoples R China Univ Wollongong Sch Comp & Informat Technol Wollongong NSW 2522 Australia

Cardiovascular diseases (CVD) seriously threaten the health of human beings, and they have caused widespread concern in recent years. At present, the diagnosis of CVD is mainly conducted by computed tomography (CT), echocardiography and nuclear magnetic resonance (NMR) technologies. NMR imaging technology is widely used in medical applications owing to its characteristics of high resolution and very low radiation. However, manual NMR image segmentation is time-consuming and error-prone, which has led to the research on automatic NMR image segmentation technologies. Researchers tend to explore the ventricular NRM image segmentation to improve the accuracy of CVD diagnosis. In this study, based on deep learning technology, we propose a layered Mask R-CNN segmentation method to segment ventricular NMR images. The experimental results show that the mean dice metrics (DM) of left ventricular segmentation and right ventricular segmentation are 0.92 and 0.89, and the Hausdorff distance (HD) metrics are 4.78 mm and 7.03 mm. Our research indicates that the proposed novel method has great potential to automate the ventricular NMR image segmentation. We also discuss the automatic abnormal ventricular systolic function detection method based on the proposed layered segmentation model.

关键词： Cardiovascular disease NMR image segmentation Layered mask R-CNN Ventricular systolic function detection

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：