发明人：陶然，周胜，董旭，范凌佳，沈千晔

申请日期：2024.04.26

摘要：本发明提供了一种基于3D打印的多分流层芯片风冷散热器及其拓扑优化方法，属于电气元件的冷却装置技术领域，用于解决现有技术中的流体与翅片之间换热不均匀、翅片的散热效率不高的问题。其中散热器包括从上到下依次设置的翅片(1)、基板(2)、导冷条(3)、芯片(4)、芯片基座(5)以及散热器底板(6)；在多个翅 ...

作者：Junhua Yang¹;Hao Huan¹;Ran Tao¹; （¹Beijing Key Laboratory of Fractional Signals and Systems, School of Information and Electronics, Beijing Institute of Technology, Beijing, China）

出处：IEEE Transactions on Aerospace and Electronic Systems 2024

关键词：Location awareness;Doppler effect;Satellites;Accuracy;Position measurement;Estimation;Apertures

摘要：High-accuracy localization of radiation sources is a prominent research area in the field of passive positioning technology. Synthetic aperture positi ...

作者：Boao Qin¹;Shou Feng^1,2;Chunhui Zhao¹;Wei Li²;Ran Tao²;Wei Xiang³; （¹College of Information and Communication Engineering and the Key Laboratory of Advanced Marine Communication and Information Technology, Ministry of Industry and Information Technology, Harbin Engineering University, Harbin, China ²School of Information and Electronics, Beijing Institute of Technology, Beijing, China ³School of Computing, Engineering and Mathematical Sciences, La Trobe University, Melbourne, VIC, Australia）

出处：IEEE Transactions on Geoscience and Remote Sensing 2024

关键词：Training;Task analysis;Hyperspectral imaging;Adaptation models;Data models;Spectral analysis;Encoding;Hyperspectral Image Classification;Few-shot Learning;Feature Disentanglement;Cross-domain Few-shot Learning;Data Sources;Feature Representation;Heterogeneous Data;Spectral Properties;Representation Learning;Spectral Domain;Target Domain;Source Characteristics;Domain Alignment;Positional Encoding;Target Scene;Low-rank Method;Classification Performance;Feature Space;Knowledge Transfer;Spectral Bands;Source Domain;Source Dataset;Domain-specific Features;Indian Pines;Pavia University;Labeled Samples;Indian Pines Dataset;Airborne Visible/Infrared Imaging Spectrometer;Projection Layer;Pixel Coordinates

摘要：Existing hyperspectral cross-domain few-shot learning (FSL) methods focus mainly on elaborating on training strategies or domain alignment algorithms, ...

作者：Maoyang Chen^1,2;Shou Feng^1,2,3;Chunhui Zhao¹;Bo Qu^2,4,5;Nan Su¹;Wei Li³;Ran Tao³; （¹College of Information and Communication Engineering and the Key Laboratory of Advanced Marine Communication and Information Technology, Ministry of Industry and Information Technology, Harbin Engineering University, Harbin, China ²Shaanxi Key Laboratory of Optical Remote Sensing and Intelligent Information Processing, Xi’an, China ³School of Information and Electronics, Beijing Institute of Technology, Beijing, China ⁴Key Laboratory of Spectral Imaging Technology CAS, Xi’an Institute of Optics and Precision Mechanics, Chinese Academy of Sciences, Xi’an, Shaanxi, China ⁵Ministry of Education Key Laboratory for Intelligent Networks and Network Security, Xi’an Jiaotong University, Xi’an, China）

出处：IEEE Transactions on Geoscience and Remote Sensing 2024

关键词：Crops;Prototypes;Hyperspectral imaging;Feature extraction;Frequency-domain analysis;Vectors;Training;Prototypical Network;Open Set Classification;Classification Accuracy;Classification Performance;Frequency Domain;Similar Spectra;Hyperspectral Image Classification;Spatial Spectral Information;Frequency Domain Information;Test Data;Transformer;Convolutional Neural Network;Convolutional Layers;K-means;Classification Results;Spectral Features;Cohen’s Kappa;Spatial Features;Spatial Dimensions;Attention Mechanism;Convolutional Neural Network Module;Salinas Dataset;Shallow Features;Features Of Different Scales;Spectral Properties;Different Levels Of Abstraction;Kinds Of Data;Ability Of The Network;Relevant Formulas;Feature Space

摘要：At present, hyperspectral image classification (HSIC) technology has been warmly concerned in all walks of life, especially in agriculture. However, e ...

发明人：胡文帅，李伟，张蒙蒙，陶然

申请日期：2024.03.21

摘要：本发明公开了一种多源遥感图像协同分类方法和系统，包括：以每个像素点为中心提取局部邻域窗口，构建训练集和测试集，并对全局信息引导向量进行初始化。联合卷积层和Transformer结构，逐层提取和更新多源遥感图像的局部和全局特征。充分利用四元数和长短时记忆网络的优势，提取有效描述多源遥感图像互补性的多模 ...

发明人：李伟，张宇翔，张蒙蒙，陶然

申请日期：2024.05.29

摘要：本发明公开了基于双向注意力机制的高光谱图像跨场景分类方法，充分考虑了跨域分类任务中单、多要素差异，提出一套完整的高光谱图像跨场景地物要素识别模型，缓解相似物种在跨时相条件下的多维度差异，可对不同地域或不同时相的高光谱影像进行粗粒度或细粒度的地物要素提取工作，极大的提高了大场景地物要素提取效率；提出的 ...

发明人：马鹏阁，李佳俊，陶然，李伟，孙俊灵，赵旭东，金秋春，李曦晴，王楠，钱金旺

申请日期：2024.05.22

摘要：本发明提供一种基于分数傅里叶变换的城市无人机目标检测方法，包括以下步骤：步骤1：高光谱目标图像反射率光谱数据库构建；步骤2：基于特征波段选择的高光谱数据降维；步骤3：基于分数傅里叶变换的城市背景无人机高光谱目标检测。本发明可以在城市复杂环境为背景下显著区分信号与背景和噪声。