检索结果-内蒙古大学图书馆

Content-Aware Distillation Network for real-time Magnetic Particle Imaging

IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT 2025年 74卷

作者： Shi, Gen Yin, Lin Bian, Zhongwei Chen, Ziwei An, Yu Hui, Hui Tian, Jie Beihang Univ Sch Engn Med Beijing 100191 Peoples R China Beihang Univ Sch Biol Sci & Med Engn Beijing 100191 Peoples R China Beihang Univ Key Lab Big Data Based Precis Med Minist Ind & Informat Technol China Beijing 100191 Peoples R China Chinese Acad Sci Inst Automat CAS Key Lab Mol Imaging Beijing 100190 Peoples R China Univ Chinese Acad Sci Sch Artificial Intelligence Beijing 100190 Peoples R China Natl Key Lab Kidney Dis Beijing 100853 Peoples R China

Magnetic particle imaging (MPI) has emerged as a promising medical imaging technique known for its high sensitivity and high imaging speed, making real-time, in vivo imaging feasible. However, existing MPI systems often require multiple repetition measurements for signal denoising. Few repetitions may result in low-quality images with increased noise, whereas many repetitions compromise temporal resolution and may introduce significant motion artifacts in dynamic imaging. Therefore, to fully exploit the advantages of MPI in real-time imaging, it is crucial to reduce the repetition number while maintaining high-quality images. In this study, we introduced a novel deep-learning (DL)-based approach, the content-aware distillation network (CAD-Net), for accelerated MPI. The method reconstructs high-quality images by denoising noisy images, typically acquired with a limited number of repetitions (tens of milliseconds). CAD-Net incorporates a proposed multiscale content-aware (MCA) block to accurately model noise distribution and enhance denoising performance. In addition, we proposed an activation-mask-based distillation strategy to reduce model processing time, particularly important for real-time imaging. Evaluation on a public real-world dataset, OpenMPI, and a simulation dataset, proved that CAD-Net outperformed existing methods in denoising performance and model efficiency. Compared to traditional methods based on multiple measurements, CAD-Net increased the frames per second (FPS) metric by approximately 70 times. Experiments on in-house data demonstrated the applicability of CAD-Net in MPI denoising in in vitro and in vivo imaging. CAD-Net improved image quality in real-time denoising with only a marginal increase in time cost. The code and data will be available at: https://***/shigen-StoneRoot/***.

关键词： image denoising knowledge distillation mag- netic particle imaging (MPI) mag- netic particle imaging (MPI) real-time real-time image denoising knowledge distillation mag- netic particle imaging (MPI) real-time

来源：评论

学校读者我要写书评

暂无评论

3D reconstruction method of welding area by fusion of coding raster and semantic segmentation network

引用

JOURNAL OF real-time image processing 2024年第5期21卷 179页

作者： Song, Limei Xu, Baolin Yang, Yangang Yuan, Jiaxing Ye, Chenchao Tiangong Univ Tianjin Key Lab Intelligent Control Elect Equipmen Tianjin 300387 Peoples R China Tianjin Univ Technol & Educ Sch Mech Engn Tianjin 300222 Peoples R China

Due to its rapid projection and capture of fringe patterns, coded structured light measurement technology is widely utilised for acquiring three-dimensional information in welding areas. To achieve real-time segmentation of the welding area for T-joint butt welding workpieces with various noise interference, this paper has developed a lightweight dual-resolution semantic segmentation network (LDRNet). This paper designed a lightweight feature extraction module that provides efficient feature representations for the network using fewer parameters and computational costs. To enhance the network's robustness to complex environmental noise, this paper introduced a multi-scale adaptive feature extraction module that can capture information at different scales of the environment. To further improve segmentation accuracy, this paper reconstructed the traditional pyramid pooling module and combined it with the CBAM attention mechanism to enhance the focus on important features. This paper proposes a local feature constraint method to improve the phase matching accuracy. Experimental results show that this paper algorithm significantly reduces Params and FLOPs by 62.02% and 65.78%, respectively, compared to the original network DDRNet-23-Slim. Moreover, it leads to an improvement in the IOU of the welding area from 74.55% to 77.43%. Additionally, this paper's proposed algorithm effectively reduces the generation of noise points through phase matching by approximately 93.88%. Consequently, this algorithm satisfactorily meets the requirements of practical production processes.

关键词： deep learning real-time semantic segmentation Epipolar constraint 3D reconstruction

来源：评论

学校读者我要写书评

暂无评论

Enhancing 3D object detection in autonomous vehicles based on synthetic virtual environment analysis

引用

image AND VISION COMPUTING 2025年 154卷

作者： Li, Vladislav Siniosoglou, Ilias Karamitsou, Thomai Lytos, Anastasios Moscholios, Ioannis D. Goudos, Sotirios K. Banerjee, Jyoti S. Sarigiannidis, Panagiotis Argyriou, Vasileios Kingston Univ Dept Networks & Digital Media Kingston upon Thames England Univ Western Macedonia Dept Elect & Comp Engn Kozani Greece MetaMind Innovat PC R&D Dept Kozani Greece Sidroco Holdings Ltd Nicosia Cyprus Univ Peloponnese Dept Informat & Telecommun Tripoli Greece Aristotle Univ Thessaloniki Phys Dept Thessaloniki Greece Bengal Inst Technol Kolkata India

Autonomous Vehicles (AVs) rely on real-time processing of natural images and videos for scene understanding and safety assurance through proactive object detection. Traditional methods have primarily focused on 2D object detection, limiting their spatial understanding. This study introduces a novel approach by leveraging 3D object detection in conjunction with augmented reality (AR) ecosystems for enhanced real-time scene analysis. Our approach pioneers the integration of a synthetic dataset, designed to simulate various environmental, lighting, and spatiotemporal conditions, to train and evaluate an AI model capable of deducing 3D bounding boxes. This dataset, with its diverse weather conditions and varying camera settings, allows us to explore detection performance in highly challenging scenarios. The proposed method also significantly improves processing times while maintaining accuracy, offering competitive results in conditions previously considered difficult for object recognition. The combination of 3D detection within the AR framework and the use of synthetic data to tackle environmental complexity marks a notable contribution to the field of AV scene analysis.

关键词： Augmented reality Object detection Scene analysis Scene understanding Object recognition deep learning Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Phase unwrapping using deep learning in holographic tomography

引用

OPTICS EXPRESS 2023年第12期31卷 18964-18992页

作者： Gontarz, Michal Dutta, Vibekananda Kujawinska, Malgorzata Krauze, Wojciech Warsaw Univ Technol Inst Micromech & Photon 8 Boboli St PL-02525 Warsaw Poland

Holographic tomography (HT) is a measurement technique that generates phase images, often containing high noise levels and irregularities. Due to the nature of phase retrieval algorithms within the HT data processing, the phase has to be unwrapped before tomographic reconstruction. Conventional algorithms lack noise robustness, reliability, speed, and possible automation. In order to address these problems, this work proposes a convolutional neural network based pipeline consisting of two steps: denoising and unwrapping. Both steps are carried out under the umbrella of a U-Net architecture;however, unwrapping is aided by introducing Attention Gates (AG) and Residual Blocks (RB) to the architecture. Through the experiments, the proposed pipeline makes possible the phase unwrapping of highly irregular, noisy, and complex experimental phase images captured in HT. This work proposes phase unwrapping carried out by segmentation with a U-Net network, that is aided by a pre-processing denoising step. It also discusses the implementation of the AGs and RBs in an ablation study. What is more, this is the first deep learning based solution that is trained solely on real images acquired with HT.

关键词： image processing Medical image processing Phase imaging Phase measurement Phase retrieval Phase unwrapping

来源：评论

学校读者我要写书评

暂无评论

Enhancing Power Lines Detection Using deep learning and Feature-Level Fusion of Infrared and Visible Light images

引用

ARABIAN JOURNAL FOR SCIENCE AND ENGINEERING 2025年第2期50卷 987-999页

作者： Aboalia, Hossam Hussein, Sherif Mahmoud, Alaaeldin Mil Tech Coll Optoelect & Automat Control Syst Dept Cairo Egypt Mil Tech Coll Comp Engn & Artificial Intelligence Dept Cairo Egypt

The detection of power lines is critical for flight safety, especially for drones and low-flying aircraft. Power line detection models help prevent collisions, reducing potential damage and preserving lives, while also safeguarding critical infrastructure. This has led to significant research efforts to develop accurate detection models. In this study, we employ paired infrared-visible power line datasets to train three distinct deep learning models. The first two models are sequential deep learning models based on VGG16 and AlexNet networks. They are tailored for detection in visible images, while they were optimized again for infrared images. For the third model, we introduce an innovative deep learning architecture utilizing Functional Application Programming Interface, affording us the flexibility to construct a multi-input model with shared layers. Our proposed model accepts paired images (visible and infrared) as inputs. Then, a feature-level fusion process is applied to merge the extracted features from both inputs and generate an enriched feature map. This approach amalgamates the advantages of visible images, which boast high resolution and rich texture features, with infrared images, which excel in high contrast and clear vision under adverse environmental conditions. Comparing the outcomes of the three models, our proposed model emerges as the front runner, boasting an impressive accuracy rate of 99.37%. Moreover, real-time processing was adopted by conducting ablation experiments to optimize the model and reduce the number of trainable parameters, resulting in an inference speed of 2.7 milliseconds per frame.

关键词： Power line detection deep learning Infrared images Visible images Fusion image classification

来源：评论

学校读者我要写书评

暂无评论

A hybrid learning frame work for recognition abnormal events intended from surveillance videos

引用

JOURNAL OF INTELLIGENT & FUZZY SYSTEMS 2023年第1期45卷 1089-1102页

作者： Babiyola, A. Aruna, S. Sumithra, S. Buvaneswari, B. Meenakshi Sundararajan Engn Coll Dept ECE Chennai Tamil Nadu India SRM Inst Sci & Technol Dept Computat Intelligence Kattankulathur Tamilnadu India JJ Coll Engn & Technol Dept ECE Trichy Tamilnadu India Panimalar Engn Coll Dept Informat Technol Chennai Tamil Nadu India

The need for a monitoring system has grown as a result of rising crime and anomalous activity. To avoid unusual incidents, the common man initiated video surveillance of important areas, which was then passed on to the government. In typical surveillance operations, surveillance devices create a vast volume of data that must be manually analysed. Manually handling huge data sets in real time results in information loss. To prevent abnormal incidents, the actions in sensitive areas can be properly monitored, evaluated, and alerted to the appropriate authorities. Previous deep learning-based activity identification methods have appeared, but the findings are inaccurate, and the proposed Hybrid Machine learning Algorithms (HMLA) incorporate two detection methods for surveillance videos like as Transfer learning (TL) and Continual learning (CL). As a result, the suspicious activity in the video may be missed. Consequently, numerous image processing and computer vision technologies were used in activity detection to decrease human effort and mistakes in surveillance operations. Activities in sensitive areas can be properly monitored and evaluated to avoid unusual incidents, and the appropriate authorities may be alerted. Hence, in order to decrease human error and effort in surveillance operations, activity recognition embraced a variety of image processing and computer vision technologies. In this present work, the capacity has constraints that impact recognition accuracy. Consequently, this research paper presents a HMLA based technique that uses feature extraction using multilayer (Long Short Term Memory) LSTM, Convolutional Neural Networks (CNN), and Temporal feature extraction using multilayer LSTM to improve identification accuracy by 96% while requiring minimal execution time. To show the superior performance of the proposed hybrid machine learning technique, a standard UCF crime dataset was utilised for experimental analysis and compared to existing deep learning

关键词： Hybrid machine learning algorithms surveillance videos transfer learning continual learning recognition abnormal events

来源：评论

学校读者我要写书评

暂无评论

A Lightweight deep-learning Algorithm for Welding Defect Detection in New Energy Vehicle Battery Current Collectors

引用

IEEE SENSORS JOURNAL 2024年第13期24卷 21655-21668页

作者： Yuan, Lei Chen, Yanrong Tang, Hai Gao, Ren Wu, Wenhuan Hubei Univ Automot Technol Sch Elect & Informat Engn Hubei Key Lab Energy Storage & Power Battery Shiyan 442002 Peoples R China

The future direction of global automotive development is electrification, and the battery current collector (BCC) is an essential component of new energy vehicle batteries. However, the welding defects in the BCC during the welding process are characterized by a disorganized distribution, extensive size variations, multiple types, and ambiguous features, posing challenges for detecting welding defects in the current collector. This article proposes a lightweight deep-learning algorithm called MGNet for detecting welding defects in the current collectors. We introduce a lightweight MDM module based on multiscale channels, which utilizes deep dynamic convolutions as its basic structure to extract compelling features while reducing computational complexity. We also propose a lightweight feature fusion network called GS_GFPN, which fully leverages the semantic information of the backbone network feature maps while reducing parameter redundancy and maintaining detection accuracy. Experimental evaluations on both the BCC surface defect database and the publicly available Northeastern University (NEU) surface defect database demonstrate that MGNet outperforms existing methods with significant improvements in detection accuracy. The mean average precision (mAP) at IoU threshold 0.5 is 93.9%-78.0%, respectively, with frames per second (FPS) of 212.8 and 238.1 and a model weight of only 3.1 M. Moreover, the algorithm is successfully deployed on the NVIDIA Jetson Nano embedded device, enabling real-time defect detection for practical industrial applications.

关键词： Batteries Welding Feature extraction Computational modeling image processing Defect detection deep learning Battery current collector (BCC) deep learning defect detection model deployment

来源：评论

学校读者我要写书评

暂无评论

A hybrid deep learning model for robust aeroengine remaining useful life prediction

引用

SIGNAL image AND VIDEO processing 2025年第7期19卷 1-18页

作者： Wan, Anping Zhang, Hua Chen, Ting AL-Bukhaiti, Khalil Wang, Wenhui Hangzhou City Univ Dept Mech Engn Hangzhou 310015 Peoples R China Zhejiang Univ Dept Mech Engn Hangzhou 310058 Peoples R China Zhejiang Univ Technol Coll Mech Engn Hangzhou 310014 Peoples R China

Accurate prediction of Remaining Useful Life (RUL) for aeroengines is critical for ensuring flight safety and optimizing maintenance. However, existing deep learning models often struggle to effectively capture multi-scale degradation patterns and long-term dependencies, particularly under varying operating conditions. To address these challenges, we propose Multi-scale Temporal-Gated Attention (MTGA), a novel hybrid deep learning framework designed to enhance RUL prediction accuracy and robustness. MTGA uniquely integrates a Multi-scale Temporal Convolutional Network (MTCN) for extracting degradation features across different temporal resolutions, a Gated Recurrent Unit (GRU) to model long-term dependencies efficiently, and an Improved Self-Attention Mechanism (ISA) to dynamically refine feature importance while suppressing noise. Unlike conventional hybrid models that apply standard attention mechanisms, MTGA's ISA enhances interpretability by adaptively weighting input features based on their relevance to degradation progression. Extensive evaluations of the Commercial Modular Aero-Propulsion System Simulation (C-MAPSS) dataset demonstrate that MTGA consistently outperforms state-of-the-art methods such as IMDSSN, PI-CNN, and ATCN. For instance, MTGA achieved RMSE reductions of 19.43% and 25.73% over IMDSSN on FD002 and FD004 datasets, respectively, and Score reductions of up to 50.17% compared to ATCN on FD004, particularly under complex operational conditions. The framework's ability to balance predictive accuracy and computational efficiency makes it well-suited for real-time predictive maintenance in aerospace applications. These findings highlight MTGA's potential to improve aviation maintenance and safety strategy decision-making.

关键词： Remaining useful life (RUL) prediction Aeroengine prognostics Multi-scale convolutional neural network Gated recurrent unit (GRU) Attention mechanism Hybrid deep learning

来源：评论

学校读者我要写书评

暂无评论

deep Idempotent Network for Efficient Single image Blind Deblurring

引用

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY 2023年第1期33卷 172-185页

作者： Mao, Yuxin Wan, Zhexiong Dai, Yuchao Yu, Xin Northwestern Polytech Univ Sch Elect & Informat Xian 710129 Shaanxi Peoples R China Univ Technol Sydney Fac Engn & Informat Technol Sydney NSW 2007 Australia

Single image blind deblurring is highly ill-posed as neither the latent sharp image nor the blur kernel is known. Even though considerable progress has been made, several major difficulties remain for blind deblurring, including the trade-off between high-performance deblurring and real-time processing. Besides, we observe that current single image blind deblurring networks cannot further improve or stabilize the performance but significantly degrades the performance when re-deblurring is repeatedly applied. This implies the limitation of these networks in modeling an ideal deblurring process. In this work, we make two contributions to tackle the above difficulties: (1) We introduce the idempotent constraint into the deblurring framework and present a deep idempotent network to achieve improved blind non-uniform deblurring performance with stable re-deblurring. (2) We propose a simple yet efficient deblurring network with lightweight encoder-decoder units and a recurrent structure that can deblur images in a progressive residual fashion. Extensive experiments on synthetic and realistic datasets prove the superiority of our proposed framework. Remarkably, our proposed network is nearly 6.5x smaller and 6.4x faster than the state-of-the-art while achieving comparable high performance.

关键词： image restoration real-time systems deep learning Mathematical models Kernel Training Convolutional neural networks Idempotent network single image blind deblurring efficient deblurring

来源：评论

学校读者我要写书评

暂无评论

S-Net: A Lightweight real-time Semantic Segmentation Network for Autonomous Driving 8th

S-Net: A Lightweight Real-Time Semantic Segmentation Network...

引用

8th International Conference on Computer Vision and image processing (CVIP)

作者： Mazhar, Saquib Atif, Nadeem Bhuyan, M. K. Ahamed, Shaik Rafi Indian Inst Technol Gauhati Assam India

ISBN: (纸本)9783031581731;9783031581748

Semantic segmentation of road-scene images for autonomous driving is a dense pixel-level prediction task performed in real-time. deep learning models make extensive efforts to improve segmentation accuracy, among which network architecture design is essential. In edge devices, this becomes more challenging due to limited computing power. While very deep encoder-decoder-based networks perform fairly accurately, their slow inference speed and many parameters make them unsuitable for small devices. Decoder-less models are fast but suffer from accuracy loss. To this end, we propose a novel architecture with a shallow decoder. We propose a building block for our network, which leverages a multi-scale feature pyramid model. The block efficiently learns semantic and contextual features based on which we design our network. It benefits from uniquely placed encoder skip connections, which are responsible for retaining low-level features to preserve boundary information, often lost in deep networks. Experiments on highly competitive Cityscapes and CamVid datasets show the efficiency of our proposed architecture. Our model gets a mean intersection-over-union score of 72.5% and 67.5% on the Cityscapes and CamVid test set, with only 0.6 Million parameters running in real-time.

关键词： Computer vision Autonomous driving Semantic segmentation deep learning

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：