检索结果-内蒙古大学图书馆

16th International conference on Signal processing systems, ICSPS 2024

作者： Yang, Gang Southwest China Institute of Electronic Technology Chengdu China

ISBN: (纸本)9781510689251

Hyperspectral image unmixing estimates a collection of constituent materials (called endmembers) and their corresponding proportions (called abundances), which is a critical preprocessing step in many remote sensing applications. In this paper, we propose a novel robust multilayer matrix factorization (RMLMF) approach for hyperspectral image unmixing. Firstly, instead of sampling all observed pixels, the endmembers are modeled as a linear combination of candidate endmembers to reduce the redundant spectra and alleviate heavy computational cost. Secondly, to explore the hidden information, the shallow decomposition is extended into multilayer structure. Furthermore, considering that hyperspectral images (HSIs) still suffer from a variety of noises, L2,1-norm is employed to measure the reconstruction error of the multilayer structure. Experimental results on both synthetic and real data sets demonstrate the superiority of the proposed RMLMF over many state-of-the-art algorithms. © 2025 SPIE.

关键词： Matrix factorization

来源：评论

学校读者我要写书评

暂无评论

A Novel Artistic image Compression Algorithm Based on Multi-Scale Adaptive Discrete Wavelet Transform

A Novel Artistic Image Compression Algorithm Based on Multi-...

引用

2025 International conference on Multi-Agent systems for Collaborative Intelligence, ICMSCI 2025

作者： Li, Yijun China Academy of Art Yuhang District Zhejiang Hangzhou311100 China

ISBN: (纸本)9798331509828

How to compress information is a challenge in the digital age, and the image compression has a wide range of application scenarios. In order to solve the challenges of low compression efficiency and high loss rate of the traditional algorithms, this study proposes a compression algorithm based on multi-scale adaptive discrete wavelet transform. The core of the proposed algorithm is to make reasonable use of the multiresolution analysis characteristics of discrete wavelet transform. Based on the feature extraction, the image information is decomposed into high-frequency and low-frequency parts. This operation can stably and effectively implement frequency domain threshold processing and reduce redundant data. The proposed algorithm optimizes the encoding efficiency of image features to the greatest extent by introducing block difference features and designing a new iterative function system. In view of the high redundancy and high complexity of the artistic images, the traditional wavelet analysis is improved and optimized, and a multi-scale adaptive model is introduced. In the experiment, the test used multiple sets of artistic image test sets and compared them with mainstream algorithms (JPEG, VAE, etc.). In PSNR analysis, the proposed algorithm achieves better performance. © 2025 IEEE.

关键词： Discrete wavelet transforms

来源：评论

学校读者我要写书评

暂无评论

Generative Waveform Design Methods in Complex Electromagnetic Spectrum Scenarios 24

Generative Waveform Design Methods in Complex Electromagneti...

引用

4th International conference on Signal processing and Communication Technology, SPCT 2024

作者： Hu, Yuhao Hu, Weidong Du, xiaoyong Lei, Gang Xi’an Institute of Electronic Engineering Shaanxi Xi’an China College of Electronic Science and Technology National University of Defense Technology Hunan Changsha China

ISBN: (纸本)9798400710636

With the development of complex spectrum environments such as 5G and 6G, the utilization of electromagnetic spectrum resources has become increasingly intensive, imposing higher demands on the performance and adaptability of traditional radar systems. The fixed waveform design approach commonly used in traditional radar systems is easily identifiable and prone to interference in complex spectrum environments, making it difficult to meet dynamically changing mission requirements. To address this challenge, this paper proposes a generative waveform design method suitable for complex spectrum environments. This method constructs a performance envelope based on radar mission requirements under multiple constraints and generates a feasible waveform set through constrained rule-based searches, thereby enhancing the flexibility and adaptability of waveform design to the environment. Experiments demonstrate that the proposed generative waveform design exhibits greater robustness and anti-interference capabilities in complex spectrum environments. Comparative analysis using classic clustering and sorting algorithms further validates the effectiveness and superiority of this method over traditional fixed waveforms, providing new insights for the design and optimization of radar systems in complex spectrum environments. © 2024 Copyright held by the owner/author(s).

关键词： Radar interference

来源：评论

学校读者我要写书评

暂无评论

Full Borehole image Data Computation Using GAN Based Model

Full Borehole Image Data Computation Using GAN Based Model

引用

2025 International Petroleum Technology conference, IPTC 2025

作者： Liu, Y. Zhou, J. Zhang, J. Li, P. Zhou, Y. China National Logging Corporation Shaanxi Xi'an China

ISBN: (纸本)9781959025436

In logging interpretation, electrical image data is used to identify important geological features such as fissures and cavities, providing crucial information for evaluating productivity. Therefore, obtaining high-quality, fully restored electrical image data is essential for both manual and AI-based image interpretation methods. However, traditional image processing algorithms often produce restored data with poor correlation between geological features across plates, as well as noise and jagged edges. To address this issue, we propose a reliable method for filling in missing sections of electrical image data that can predict high-quality images using deep learning algorithms. While traditional image inpainting algorithms fail to extract meaningful features from pad data, deep learning networks using the encoder-decoder architecture can automatically learn representations of borehole images. In this work, we propose a Generative Adversarial Network (GAN) based inpainting model that generates high-quality missing pad data with semantically meaningful features as a reference for image logging interpretation. We achieved a U-Net based model as an encoder to aggregate contextual information from various feature maps through borehole image data. Based on the encoder, a simple up-sampling decoder with convolution layers is applied to predict the data in the missing area of borehole image data. The pixel-wise discriminator is used to distinguish between real image data and the corresponding synthesized patch. Our proposed model outperforms traditional inpainting methods, achieving an average accuracy of over 89% when inpainting missing areas of real borehole image data, and further validating the superiority of our method through analyzing the image log data with different azimuths. As a result, the encoder in this work also demonstrates its ability to be an efficient backbone model for extracting significant information from borehole image data. Copyright 2025, Internat

关键词： Electric logging

来源：评论

学校读者我要写书评

暂无评论

Enhancing domain adaptation for plant diseases detection through Masked image Consistency in Multi-Granularity Alignment

引用

EXPERT systems WITH APPLICATIONS 2025年 276卷

作者： Guo, Guinan Lai, Songning Wu, Qingyang Shou, Yuntao Shi, Wenxu Sun Yat Sen Univ Sch Geog & Planning Guangzhou 510700 Peoples R China Shandong Univ Sch Informat Sci & Engn Qingdao 266237 Peoples R China Univ Calif Los Angeles Dept Environm Hlth Sci Los Angeles CA 90095 USA Xi An Jiao Tong Univ Sch Comp Sci Xian 710049 Peoples R China Univ Chinese Acad Sci Aerosp Informat Res Inst Chinese Acad Sci Beijing 100049 Peoples R China

Plant diseases have various impacts on the environment, encompassing ecosystems, water bodies, soils, and biodiversity. The most direct adverse effect of plant diseases on humans lies in their potential to reduce crop yields, posing a threat to food security. Employing deep learning models for object detection of plant diseases can effectively enhance diseases prevention and control, thereby exerting a positive influence on both ecological environment and sustainable agricultural development. However, during the detection of plant diseases, disparities in environmental conditions, locations, and data styles during data collection can result in significant variations in feature distributions between the training and testing datasets of deep learning models. In reality, the presence of such inter-domain differences leads to a notable decrease in the model's detection performance. Therefore, the primary objective of our study is to enhance the domain generalization capability of plant diseases object detection models. We employed an enhanced version of the object detector and introduced the use of Multi-Granularity Alignment (MGA) for domain adaptation, coupled with Masked image Consistency (MIC) for domain augmentation (MIC-MGA). Initially, we refined and optimized the backbone and neck sections of the object detector, and then combined it with the Multi-Granularity Alignment domain adaptation module, and finally used the Masked image Consistency module inspired by natural language processing to bolster the model's domain generalization detection capabilities. To further scrutinize the cross- domain detection robustness of MIC-MGA, we innovatively applied image style transformation techniques for data augmentation. Finally, we validated the model's performance using plant diseases data collected from various scenarios. Our approach outperformed other algorithms in all three tasks, achieving the highest mean Average Precision (mAP). Our research results affirm that the

关键词： Plant diseases object detection Sustainable development Cross domain Domain adaptation Deep learning

来源：评论

学校读者我要写书评

暂无评论

Multi-modal Cooperative Perception of Constrained Multi-UAV Platform 17th

Multi-modal Cooperative Perception of Constrained Multi-UAV ...

引用

17th International conference on Intelligent Robotics and Applications

作者： Zhang, Yiheng Cui, Yuanzhe Su, Yongbo Li, Tong Wang, Bingzheng Tang, Qirong Tongji Univ Lab Robot & Multibody Syst Sch Mech Engn Shanghai Peoples R China

ISBN: (纸本)9789819607884;9789819607891

The constrained multi-UAV platform is composed of multiple unmanned aerial vehicles (UAVs) connected physically and mutually. When multiple UAVs operate collaboratively on an interconnected platform, challenges in external environment perception arise due to limited movement range and sensor disturbance among UAVs. Additionally, real-time perception is crucial for various UAV tasks, posing a significant challenge for conventional single-modal algorithms to balance speed and accuracy effectively. This paper proposes an innovative multi-modal collaborative sensing method for multi-UAV systems, integrating point cloud and image processing algorithms. A feature-level fusion is employed by our framework to accurately determine target positioning in real-time. The multi-UAV decision layer utilizes a Kalman Filter to extract data from multiple individual outputs, yielding stable and accurate positioning and classification information. Simulation experiments validate the robustness of this method, reducing disturbance from structural and mutual obstructions among UAVs, and providing accurate target positioning.

关键词： Multi-UAV Perception Sensor Fusion Target Detection

来源：评论

学校读者我要写书评

暂无评论

Hardware Implementation of image processing Morphological and Convolution Operations as SoC on FPGAs 24

Hardware Implementation of Image Processing Morphological an...

引用

7th International conference on Digital Medicine and image processing, DMIP 2024

作者： Moussa, Alfred M. Groves, Richard Rafla, Nader Electrical and Computer Engineering Department Boise State University BoiseID United States

ISBN: (纸本)9798400709586

Efficient image processing architectures are consistently in demand across a multitude of applications, particularly those customized for resource-constrained systems-on-chip (SoC). The increasing need for high-performance image processing in various sectors has driven the development of specialized architectures. However, deploying such architectures on platforms with limited resources, such as SoCs, poses significant challenges. Furthermore, the implementation of complex algorithms to handle large datasets using software solutions often leads to slower response times, prompting exploration into hardware implementations. Field-Programmable Gate Arrays (FPGAs) are becoming popular for hardware implementations because of their attributes: low latency, connectivity, parallel computing capabilities, and flexibility. Consequently, the utilization of FPGA-based implementations has resulted in faster and more efficient performance of unique architectures tailored to specific requirements. This paper presents a novel hardware/software co-design approach to implement erosion, dilation, and neighborhood image processing operations on the FPGA development board, "Zedboard". In this approach, the FPGA is programmed by connecting it to a PC via USB, facilitating the transfer of an image pixel by pixel. The pixels are temporarily stored in on-chip DDR and accessed through DMA (Direct Memory Access) until they are requested by an interrupt signal from the image processing IP, at which point they are moved to line buffers for faster processing. Once processed, the image is transmitted back to the PC via UART, facilitating pixel-by-pixel transfer for verification, where it is compared with a reference image generated using Python. This comparison confirms a 99.22% match between the processed image and the reference image, with the discrepancy occurring at the image's edges due to initial padding. Additionally, the time required to process the entire image was measured and displayed

关键词： System-on-chip

来源：评论

学校读者我要写书评

暂无评论

Visual inspection of coatings in hydrogen fuel cells 24

Visual inspection of coatings in hydrogen fuel cells

引用

3rd International conference on Artificial Intelligence and Intelligent Information processing, AIIIP 2024

作者： Li, Chenjie Zhou, Mian Pan, Weiwei Tianjin University of Technology Tianjin China Xi’an Jiaotong-Liverpool University Jiangsu Suzhou China

ISBN: (纸本)9798400707308

In industrial production, the presence or absence of defects on the surface of industrial products will have a great impact on the quality of some products. In this paper, we propose a method that combines traditional image recognition and deep learning-based object detection algorithms, which aims to detect whether there are defects on the coating of hydrogen fuel cells and identify the types and sizes of defects. This method integrates multiple functions for detecting different defects, such as the overall detection and identification of defects such as large area missing coating, uneven application, empty bands and dimensions that do not meet the standard, and detailed detection such as the detection of light transmission points and the detection of light-sladen areas of the coating material (referred to as dark layers in this paper). The system can be installed directly on the production line to detect defects in the coating material of hydrogen fuel cells. © 2024 Copyright held by the owner/author(s).

关键词： Coated material

来源：评论

学校读者我要写书评

暂无评论

Glare-SNet: Unsupervised Glare Suppression Balance Network 27th

Glare-SNet: Unsupervised Glare Suppression Balance Network

引用

27th International conference on Pattern Recognition, ICPR 2024

作者： Li, Pei Zuo, Chengyu Wei, Wangjuan Pan, xiaoying Wang, Zhanhao School of Computer Science and Technology Xi’an University of Posts and Telecommunications Xi’an710121 China Shaanxi Key Laboratory of Network Data Analysis and Intelligent Processing Xi’an China Shaanxi Key Laboratory of Intelligent Media Computing and interaction Xi’an China Xi’an Key Laboratory of Big Data and Intelligent Computing Xi’an China

ISBN: (纸本)9783031781681

In light of the problems associated with glare and halo effects in low-light images, as well as the inadequacy of existing processing algorithms in handling details, a glare suppression balance network based on unsupervised learning has been designed and implemented in this study. A network that perceives brightness through light decomposition has been proposed, and an illumination learning network has been designed. The idea of integration has been employed to form an unsupervised glare suppression balance network, aiming to tackle the issues of glare and the accompanying halo effects. Through objective and subjective evaluations of real images, it has been proven that the method introduced in this paper surpasses other methods in suppressing glare and enhancing the overall quality of low-light images. Furthermore, it has been shown to exhibit good performance and effectiveness in target detection. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： image enhancement

来源：评论

学校读者我要写书评

暂无评论

processing and Classification algorithms of Medical images 6

Processing and Classification Algorithms of Medical Images

引用

6th IEEE International conference on image processing, Applications and systems, IPAS 2025

作者： Seitaj, Oltiana Mechanics University of Roma Tre Department of Industrial Engineering Electronics Rome Italy

ISBN: (纸本)9798331506520

This paper evaluates the impact of hybrid deep learning approaches on lung tumor segmentation by combining traditional image processing techniques with advanced AI-driven models. The study integrates Convolutional Neural Networks (CNNs) with preprocessing methods such as noise reduction, adaptive thresholding, and contrast enhancement to address challenges associated with complex anatomical structures and variability in medical image quality. A novel hybrid framework is proposed, leveraging traditional methods to preprocess data and improve input quality for deep learning models, ultimately enhancing segmentation accuracy and *** effectiveness of the proposed approach is assessed using quantitative performance metrics, including Dice Similarity Coefficient (DSC), Hausdorff Distance, Jaccard Index, Precision, and Recall. Preliminary results indicate significant improvements in tumor boundary detection and reduced false-positive rates compared to existing methods. By streamlining segmentation workflows and enabling near-realtime applications in clinical settings, this research offers a pathway to improved diagnostic accuracy, treatment planning, and workflow *** implications include the potential for integration into clinical imaging pipelines, fostering advancements in computer-assisted diagnosis and personalized treatment strategies. This study underscores the value of hybrid methodologies in addressing current limitations and paving the way for more precise and efficient medical image segmentation. © 2025 IEEE.

关键词： Convolutional neural networks

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：