检索结果-内蒙古大学图书馆

DES to HSC: Detecting low-surface-brightness galaxies in the Abell 194 cluster using transfer learning

ASTRONOMY & ASTROPHYSICS 2025年 695卷

作者： Thuruthipilly, H. Junais, J. Koda, J. Pollo, A. Yagi, M. Yamanoi, H. Komiyama, Y. Romano, M. Malek, K. Donevski, D. Natl Ctr Nucl Res Pasteura 7 Warsaw Poland Inst Astrofis Canarias Via Lactea S-N San Cristobal la Laguna 38205 Spain Univ La Laguna Dept Astrofis San Cristobal la Laguna 38206 Spain SUNY Stony Brook Dept Phys & Astron Stony Brook NY 11794 USA Jagiellonian Univ Astron Observ Orla 171 Krakow Poland Natl Astron Observ Japan Mitaka Tokyo 1818588 Japan Seikei Univ Fac Sci & Technol 3-3-1 Kichijoji Kitamachi Musashino Tokyo 1808633 Japan Hosei Univ Fac Sci & Engn Dept Adv Sci 3-7-2 Kajino Cho Koganei Tokyo 1848584 Japan Grad Univ Adv Studies SOKENDAI Mitaka Tokyo 1818588 Japan Max Planck Inst Radioastron Auf dem Hugel 69 DE-53121 Bonn Germany INAF Osservatorio Astron Padova Vicolo Osservatorio 5 I-35122 Padua Italy Aix Marseille Univ CNRS CNES LAM Marseille France

Context. Low-surface-brightness galaxies (LSBGs) are important for understanding galaxy evolution and cosmological models. Nevertheless, the physical properties of these objects remain unknown, as even the detection of LSBGs can be challenging. Upcoming large-scale surveys are expected to uncover a large number of LSBGs, which will require accurate automated or machine learningbased methods for their detection. Aims. We study the scope of transfer learning for the identification of LSBGs. We used transformer models trained on Dark Energy Survey (DES) data to identify LSBGs from dedicated Hyper Suprime-Cam (HSC) observations of the Abell 194 cluster, which are two magnitudes deeper than DES. A new sample of LSBGs and ultra-diffuse galaxies (UDGs) around Abell 194 was compiled, and their properties were investigated. Methods. We used eight models, divided into two categories: LSBG Detection Transformer (LSBG DETR) and LSBG vision Transformer (LSBG ViT). The data from DES and HSC were standardised based on the pixel-level surface brightness. We used an ensemble of four LSBG DETR models and another ensemble of four LSBG ViT models to detect LSBGs. This was followed by a singlecomponent S & eacute;rsic model fit and a final visual inspection to filter out potential false positives and improve sample purity. Results. We present a sample of 171 LSBGs in the Abell 194 cluster using HSC data, including 87 new discoveries. Of these, 159 were identified using transformer models, and 12 additional LSBGs were found through visual inspection. The transformer model achieves a true positive rate of 93% in HSC data without any fine-tuning. Among the LSBGs, 28 were classified as UDGs. The number of UDGs and the radial UDG number density suggests a linear relationship between UDG numbers and cluster mass on a log scale. The UDGs share similar S & eacute;rsic parameters with dwarf galaxies and occupy the extended end of the R-eff - M-g plane, suggesting they might be an extended sub-po

关键词： methods: data analysis methods: observational techniques: image processing catalogs galaxies: dwarf galaxies: evolution

来源：评论

学校读者我要写书评

暂无评论

Prediction of Cervical Cancer Staging via Diffusion-Weighted MR Radiomic Features Analysis

Prediction of Cervical Cancer Staging via Diffusion-Weighted...

引用

Artificial Intelligence, applications, Innovation and Ethics (AI2E), International Conference for

作者： Zarina Ramlia Sri Herwiningsih Mohd Saiful Asmal Mohd Amiruddin Zaharudin Haron Muhammad Khalis Abdul Karim Department of Physics Universiti Putra Malaysia Selangor Malaysia Department of Physics Universitas Brawijaya Malang Indonesia Department of Radiology Institut Kanser Negara Putrajaya Malaysia

ISBN: (数字)9798331513733

ISBN: (纸本)9798331513740

Contemporary technology has made it possible to diagnose diseases with greater precision, enabling doctors to determine tumor stages by measuring and analyzing features in medical images. This study aimed evaluate radiomic features of DWI-MRI cervical and to classify tumors based on their stages using several machine learning models. The study involved 60 patients from the National Cancer Institute in Putrajaya using which 852 radiomic were derived from each patient. The patients were divided into two groups based on the stage of cervical cancer: Group one is defined by stages I and ii, while group two is labelled by stages iiI and IV. A semi-automatic process was used to segment the image data and extract various features such as shape, wavelength, and both first and second-order statistics (GLSZM) from the segmented images. In our study, we used three different model algorithms: Logistic Regression (LR), Support Vector machine (SVM) and Decision Tree (DT). We conducted a detailed analysis to compare their performance. The SVM algorithm stood out as the top performer, achieving the highest accuracy among the models tested. We evaluated its effectiveness using multiple performance metrics, which confirmed its superiority in processing features extracted from enhanced images. Notably, the SVM classifier demonstrated excellent results, with an accuracy of 0.77, and precision of 0.63. It also showed a robust area under the ROC curve of 96%, highlighting its effectiveness in classification tasks. Furthermore, this research emphasizes the significance of image enhancement as the useful preprocessing step which increases the classification accuracy.

关键词： Support vector machines image segmentation Technological innovation Accuracy Feature extraction Decision trees image enhancement Cervical cancer Radiomics Tumors

来源：评论

学校读者我要写书评

暂无评论

Artificial Intelligence-Powered Internet of Things for Farming and Plant Research

Artificial Intelligence-Powered Internet of Things for Farmi...

引用

Electronics and Renewable Systems (ICEARS), International Conference on

作者： A. Padma P. Suresh G. Narsimlu M Amarnath B. R Sreedhar Department of Mathematics Chaitanya Bharathi Institute of Technology

ISBN: (数字)9798331509675

ISBN: (纸本)9798331509682

The wearable fiber optic sensors for plants to monitor microenvironment and growth wearable fiber optic plant sensors for microclimate and growth monitoring. The plant agricultural sectors are facing challenges because to population expansion and climate ***-powered computer vision-based plant growth monitoring system the goal of this research project is to use computer vision to create an AI-based plant growth monitoring system. The technology will allow for automatic and real-time evaluation of plant growth and health by utilizing computer vision algorithms and artificial intelligence approaches. Conventional techniques for tracking plant growth require a lot of work and time. By employing image processing techniques to analyze plant photos and extract pertinent features leaf area, plant height, and width, the suggested solution will overcome these drawbacks. Using the collected information, machine learning methods will be used to train models that can identify patterns and forecast plant development. The system will be made to adapt to various plant species and environmental circumstances. This AI-based plant growth monitoring system will help farmers, researchers, and plant enthusiasts by offering precise and timely insights into plant health and growth. This will allow them to make well-informed decisions about plant care and optimize growth conditions. By providing an economical, scalable, and effective system with numerous applications in horticulture, agriculture, and environmental research, this research project seeks to transform plant growth monitoring.

关键词： Temperature sensors Optical fiber sensors Computer vision Plants (biology) Feature extraction Internet of Things Biomedical monitoring Monitoring Intelligent sensors Wearable sensors

来源：评论

学校读者我要写书评

暂无评论

Data Fusion Approach for Unmodified UAV Tracking with vision and mmWave Radar

Data Fusion Approach for Unmodified UAV Tracking with Vision...

引用

International Conference on Unmanned Aircraft Systems (ICUAS)

作者： Guilherme Amaral João J. Martins Pedro Martins André Dias José Almeida Eduardo Silva INESC TEC - Institute for Systems and Computer Engineering Technology and Science Porto Portugal ISEP Polytechnic of Porto rua Dr. António Bernardino de Almeida Porto Portugal

ISBN: (数字)9798331513283

ISBN: (纸本)9798331513290

The knowledge of the precise 3D position of a target in tracking applications is a fundamental requirement. The lack of a low-cost single sensor capable of providing the three-dimensional position (of a target) makes it necessary to use complementary sensors together. This research presents a Local Positioning System (LPS) for outdoor scenarios, based on a data fusion approach for unmodified UAV tracking, combining a vision sensor and mmWave radar. The proposed solution takes advantage of the radar's depth observation ability and the potential of a neural network for image processing. We have evaluated five data association approaches for radar data cluttered to get a reliable set of radar observations. The results demonstrated that the estimated target position is close to an exogenous ground truth obtained from a Visual Inertial Odometry (VIO) algorithm executed onboard the target UAV. Moreover, the developed system's architecture is prepared to be scalable, allowing the addition of other observation stations. It will increase the accuracy of the estimation and extend the actuation area. To the best of our knowledge, this is the first work that uses a mmWave radar combined with a camera and a machine learning algorithm to track a UAV in an outdoor scenario.

关键词： Visualization Target tracking Radar clutter Airborne radar Data integration Radar imaging vision sensors Radar tracking Autonomous aerial vehicles Millimeter wave communication

来源：评论

学校读者我要写书评

暂无评论

Proceedings of SPIE-image processing: machine vision applications

Proceedings of SPIE-Image Processing: Machine Vision Applica...

引用

image processing: machine vision applications

ISBN: (纸本)9780819469854

The proceedings contain 34 papers. The topics discussed include: vision robot with rotational camera for searching ID tags;a unifying software architecture for model-based visual tracking;video object tracking using chamber matching and condensation particle filter;compact and thin multi-lens system for machine vision applications;video-rate or high-precision: a flexible range imaging camera;polarization imaging for industrial inspection;machine vision approach for improving accuracy of focus-based depth measurements;3D geometric modelling of hand-woven textile;a novel circle detection method using Randon transform;an algorithm for automated registration of maps and images based on feature detection and mutual information;automatic cell segmentation and classification using morphological features and Bayesian networks;multi-model geometrical fitting for wide baseline image matching;and constraint optimization method for line fitting.

关键词： image segmentation

来源：评论

学校读者我要写书评

暂无评论

Novel computational workflows for natural and biomedical image processing based on hypercomplex algebras

arXiv

引用

arXiv 2025年

作者： Valous, Nektarios A. Hitzer, Eckhard Duşe, Dragoş Moraleda, Rodrigo Rojas Popp, Ferdinand Suarez-Carmona, Meggy Berthel, Anna Papageorgiou, Ismini Fremd, Carlo Rölle, Alexander Westhoff, Christina C. Lenoir, Bénédicte Halama, Niels Zörnig, Inka Jäger, Dirk Im Neuenheimer Feld 460 Heidelberg69120 Germany Im Neuenheimer Feld 672 Heidelberg69120 Germany Heidelberg University Im Neuenheimer Feld 267 Heidelberg69120 Germany College of Liberal Arts International Christian University Osawa 3-10-2 Mitaka Tokyo181-8585 Japan Im Neuenheimer Feld 460 Heidelberg69120 Germany Synaptiq Str. Clinicilor 23 Cluj-Napoca400006 Romania Im Neuenheimer Feld 280 Heidelberg69120 Germany Im Neuenheimer Feld 280 Heidelberg69120 Germany Obere Zahlbacher Straße 63 Building 911 Mainz55131 Germany Institute of Radiology Südharz Hospital Nordhausen Dr.-Robert-Koch-Straße 39 Nordhausen99734 Germany Institute of Diagnostic and Interventional Radiology Jena University Hospital Am Klinikum 1 Jena07747 Germany Im Neuenheimer Feld 460 Heidelberg69120 Germany Im Neuenheimer Feld 280 Heidelberg69120 Germany Baldingerstrasse Marburg35033 Germany Im Neuenheimer Feld 280 Heidelberg69120 Germany University Medical Center of Johannes Gutenberg University Mainz Langenbeckstrße 1 Mainz55131 Germany Department of Hematology and Medical Oncology III. Medical Clinic and Polyclinic University Medical Center of Johannes Gutenberg University Mainz Langenbeckstrße 1 Mainz55131 Germany

Hypercomplex image processing extends conventional techniques in a unified paradigm encompassing algebraic and geometric principles. This work leverages quaternions and the two-dimensional orthogonal planes split framework (splitting of a quaternion - representing a pixel - into pairs of orthogonal two-dimensional planes) for natural and biomedical image analysis through the following computational workflows and outcomes: natural and biomedical image re-colorization, natural image de-colorization, natural and biomedical image contrast enhancement, computational re-staining and stain separation in histological images, and performance gains in machine/deep learning pipelines for histological images. The workflows are analyzed separately for natural and biomedical images to showcase the effectiveness of the proposed approaches in each instance. The proposed workflows can regulate color appearance (e.g. with alternative renditions and grayscale conversion) and image contrast, be part of automated image processing pipelines (e.g. isolating stain components, boosting learning models), and assist in digital pathology applications (e.g. enhancing biomarker visibility, enabling colorblind-friendly renditions). Employing only basic arithmetic and matrix operations, this work offers a computationally accessible methodology - in the hypercomplex domain - that showcases versatility and consistency across image processing tasks and a range of computer vision and biomedical applications. Furthermore, the proposed non-data-driven methods achieve comparable or better results (particularly in cases involving well-known methods) to those reported in the literature, showcasing the potential of robust theoretical frameworks with practical effectiveness. Results, methods, and limitations are detailed alongside discussion of promising extensions, emphasizing the notable potential of feature-rich mathematical/computational frameworks for natural and biomedical images. © 2025, CC BY-NC-ND.

关键词： Medical image processing

来源：评论

学校读者我要写书评

暂无评论

LEGO: LLM-enhanced genetic optimization for underwater robot image restoration

引用

Pattern Recognition 2025年 168卷

作者： Ma, Yuanzheng Xia, Fan Lin, Liwei Guan, Xun Tsinghua Shenzhen International Graduate School Tsinghua University Shenzhen Guangdong 518055 China Department of Mechanical Engineering University of California Berkeley Berkeley 94709 CA United States

The absence of ground-truth labels in deep-sea image processing makes it impossible to carry out end-to-end training steps for typical machine learning schemes. Here, an agent-based genetic algorithm is successfully developed by assembling a pre-trained large language model (LLM) to restore underwater images. The LLM-enhanced genetic optimization (LEGO) can evaluate the image restoration quality by analyzing descriptive cues rather than relying on high-resolution references. Experimental results demonstrate an 80.9% improvement in the image clarity and color balance, a 13.87% increase in the F1 score, and a 21.25% increase in the recall score for YOLO-based object detection in unexplored waters. Additional tests on feature matching and semantic segmentation validate the versatility of the proposed approach to significantly enhance the robot's vision under varying water conditions across multiple applications, even in the absence of high-resolution images. © 2025 Elsevier Ltd

关键词： Genetic algorithm image restoration Large language model Robot vision Underwater robot

来源：评论

学校读者我要写书评

暂无评论

An Innovative Decision Support Glucose Metabolism Disorder Prediction Approach Based on Norm-Learning and Pre-Trained Architecture Pattern

An Innovative Decision Support Glucose Metabolism Disorder P...

引用

International Conference on Inventive Computation Technologies (ICICT)

作者： PreethaRajagopalan. M Anuratha. V Elamparithi. M Department of Computer Science Kamalam College of Arts and Science Anthiyur Bharathiar University Coimbatore Tamil Nadu India

ISBN: (数字)9798331512248

ISBN: (纸本)9798331512255

Glucose metabolism disorder has the potential to adversely affect various organs, including the eyes, coronary arteries, and kidneys, through multiple mechanisms. Consequently, it ranks among the foremost contributors to disability and premature mortality globally. The number of deaths from glucose metabolism disorder is increasing year by year, so developing a system that can correctly detect people with glucose metabolism disorder is essential. This study presents an advanced medical decision-making system that uses L-norm Support Vector machine with pre-trained UNet (ι - SVM with UNet) to forecast glucose metabolism disorder. These state-of-the-art algorithms, well-known in computer vision, image analysis, and natural language processing, show promise in producing highly accurate outcomes when used for predictive and medical diagnostic applications. Furthermore, their integration with medical data can significantly improve the decision-making process's adaptability, transparency, and effectiveness. A comparative analysis is conducted between the ι - SVM with UNet algorithm's performance, various established methodologies, and several widely used machine learning (ML) techniques. The findings of the study show that the suggested approach, which is based on ι - SVM with UNet technology, performs admirably, attaining an F1-score of 96% and an accuracy of 96%. This enhancement can increase the correctness of the final judgment while saving effort, labor, and time in healthcare services.

关键词： Support vector machines Representation learning Accuracy Decision making Mortality Prediction algorithms Natural language processing Glucose Metabolism Medical diagnostic imaging

来源：评论

学校读者我要写书评

暂无评论

Intelligence Pest Detection and Control in Agriculture using Computer vision and Deep Learning

Intelligence Pest Detection and Control in Agriculture using...

引用

International Conference on ent Computing and Control Systems (ICICCS)

作者： S. Ashok Kumar Fahimuddin Shaik B. Yashwitha P. Vidyavathi A.Uma Maheswari P. Spurthi Department of AI&DS Annamacharya University Andhra Pradesh Department of ECE Annamacharya University Andhra Pradesh Department of AI&DS Annamacharya Institute of Technology and Sciences Andhra Pradesh

ISBN: (数字)9798331512088

ISBN: (纸本)9798331512095

One of the most significant challenges in agriculture over the years has been the detection and also control of pests and while it continues to this day, the advent of technology in the agricultural sector is providing solutions that are incredibly promising. Pests cause significant worldwide crop loss and represent a threat to food security and farmer livelihoods. Conventional pest management approaches revolve around labour-intensive strategies, large applications of pesticides, and reactionary measures, resulting in environmental degradation and sub-optimal crop yields. This work utilizes the power of AI(Artificial Intelligence),computer vision and deep learning to create a smart pest detection system. It treats pest identification and damage detection as real-time visual data processing tasks with machine learning approaches, making it possible to swiftly and accurately locate pests and their damage in the photos taken through the camera, facilitating extra prompt preventive measures. Working under the hood, the algorithms adapt to changing environmental conditions and shifting pest behaviours, learning as they go. Most of the pest detection systems today are manual, requiring manual visual inspection and image recognition approaches, this solution not only provides the next level of system but it also paves the way for pest management in the most efficient, accurate & sustainable way.

关键词： Deep learning Computer vision Visualization Accuracy Pesticides Manuals Control systems Real-time systems Agriculture Farming

来源：评论

学校读者我要写书评

暂无评论

Automatic generation of input-aware approximate arithmetic circuits

Automatic generation of input-aware approximate arithmetic c...

引用

IEEE Design and Diagnostics of Electronic Circuits and Systems (DDECS)

作者： Mario Barbareschi Salvatore Barone Alberto Bosio Bastien Deveautour Ali Piri Marcello Traiola Department of Electrical Engineering and Information Technologies University of Naples Federico II Naples Italy Department of Computer Science and Technologies Pegaso Digital University Naples Italy CNRS UCBL CPE Lyon INL UMR5270 Univ Lyon ECL INSA Lyon Ecully France CNRS IETR-UMR 6164 Nantes Université Nantes France CNRS Inria IRISA - UMR 6074 Univ Rennes Rennes France

ISBN: (数字)9798331528010

ISBN: (纸本)9798331528027

Approximate Computing (AxC) is systematically applied across various abstraction levels to reduce overheads and enhance the performance of applications such as image processing and machine learning. However, AxC does not typically consider the specific workload (i.e., data input) of a given application. For instance, in signal processing applications like filters, some inputs are constants (filter coefficients), which allows for an additional level of approximation by considering the specific input distribution. This method is known as "Input-Aware Approximation" (IAA) and has shown potential advantages in previous studies. Unfortunately, existing input-aware design methodologies lack scalability as they mostly depend on ad-hoc, non-automatic design approaches, limiting their applicability. In this paper, we investigate how the input-aware approximate design approach can be integrated into a systematic, generic, and automatic design flow. We employ state-of-the-art approximation and multi-objective optimization techniques to achieve inputawareness. Our experimental results, focusing on classical signal processing applications like FIR filters, demonstrate that the input-aware approach can provide significant savings in both area and power consumption.

关键词： Systematics Filters Power demand Philosophical considerations Finite impulse response filters Scalability Approximate computing Hardware Optimization Arithmetic

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：