检索结果-南通市图书馆

Parallelizing maximum likelihood classification on computer cluster and graphics processing unit for supervised image classification

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

International Journal of Digital Earth 2017年第7期10卷 737-748页

作者： Xuan Shi bowei xue Department of Geosciences University of ArkansasFayettevilleARUSA

Supervised image classification has been widely utilized in a variety of remote sensing *** large volume of satellite imagery data and aerial photos are increasingly available,high-performance image processing solutions are required to handle large scale of *** paper introduces how maximum likelihood classification approach is parallelized for implementation on a computer cluster and a graphics processing unit to achieve high performance when processing big imagery *** solution is scalable and satisfies the need of change detection,object identification,and exploratory analysis on large-scale high-resolution imagery data in remote sensing applications.

关键词： Maximum likelihood classification supervised classification parallel computing graphics processing unit

Detecting events from the social media through exemplar-enhanced supervised learning

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

International Journal of Digital Earth 2019年第9期12卷 1083-1097页

作者： Xuan Shi bowei xue Ming-Hsiang Tsou Xinyue Ye Brian Spitzberg Jean Mark Gawron Heather Corliss Jay Lee Ruoming Jin Department of Geosciences University of ArkansasFayettevilleARUSA b Department of Geography San Diego State UniversitySan DiegoCAUSA Urban Informatics&Spatial Computing Lab Department of InformaticsNew Jersey Institute of TechnologyNewarkNJUSA School of Communication San Diego State UniversitySan DiegoCAUSA Department of Linguistics San Diego State UniversitySan DiegoCAUSA School of Public Health San Diego State UniversitySan DiegoCAUSA College of Environment and Planning Henan UniversityKaifengPeople’s Republic of China Department of Geography Kent State UniversityKentOHUSA Department of Computer Sciences Kent State UniversityKentOHUSA

Understanding and detecting the intended meaning in social media is challenging because social media messages contain varieties of noise and chaos that are irrelevant to the themes of *** example,conventional supervised classification approaches would produce inconsistent solutions to detecting and clarifying whether any given Twitter message is really about a wildfire ***,a renovated workflow was designed and *** workflow consists of four sequential procedures:(1)Apply the latent semantic analysis and cosine similarity calculation to examine the similarity between Twitter messages;(2)Apply Affinity Propagation to identify exemplars of Twitter messages;(3)Apply the cosine similarity calculation again to automatically match the exemplars to known training results,and(4)Apply accumulative exemplars to classify Twitter messages using a support vector machine *** overall correction ratio was over 90%when a series of ongoing and historical wildfire events were examined.

关键词： Social media Twitter wildfire supervised learning

改进的DeepLabV3+模型用于震后高分遥感影像滑坡识别

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

遥感学报 2024年第9期28卷 2293-2305页

作者：赵通张双成何晓宁薛博维查富康长安大学地质工程与测绘学院西安710054 地理信息工程国家重点实验室西安710054 西部矿产资源与地质工程教育部重点实验室西安710054 中科星图空间技术有限公司西安710054 中国科学院空天信息创新研究院北京100094

利用“深度学习+遥感”在高分遥感影像上快速识别滑坡,对于灾后应急救援工作的开展具有非常重要的现实意义。但目前为止,公开的深度学习滑坡识别数据集较少,难以满足相关研究人员利用深度学习方法开展滑坡识别研究的任务需求;基于此,本... 详细信息

利用“深度学习+遥感”在高分遥感影像上快速识别滑坡,对于灾后应急救援工作的开展具有非常重要的现实意义。但目前为止,公开的深度学习滑坡识别数据集较少,难以满足相关研究人员利用深度学习方法开展滑坡识别研究的任务需求;基于此,本文使用谷歌地球影像和高分六号遥感影像制作并公开了一个样本数量达到10000+,空间分辨率为2 m的深度学习滑坡数据集,用于相关学者开展研究。此外,为有效识别滑坡,提取滑坡边界及细节等信息,本文提出了一种改进的DeepLabV3+模型用于滑坡识别,即在DeepLabV3+模型的基础上,引入通道注意力机制特征融合模块和转置卷积来重构模型。实验结果表明,与主流算法(FCN,U-Net,SegNet,DeepLabV3+)相比,改进的DeepLabV3+模型在滑坡的边界及细节方面提取效果更好,所得结果更接近于真实标签。本文的相关代码和数据可在(https://***/ZhaoTong0203/landslides_identification_model_***)获得。

关键词：高分遥感影像滑坡数据集深度学习 DeepLabV3+ 高分六号

在线全文

学校读者我要写书评

暂无评论

一阶全卷积遥感影像倾斜目标检测

中国图象图形学报 2022年第8期27卷 2537-2548页

作者：周院杨庆庆马强薛博维孔祥楠中科星图空间技术有限公司西安710000

目的主流深度学习的目标检测技术对自然影像的识别精度依赖于锚框设置的好坏,并使用平行于坐标轴的正框表示物体位置,而遥感影像中地物目标具有尺寸多变、分布密集、长宽比悬殊且朝向不定的特点,更宜通过与物体朝向一致的斜框表示其位... 详细信息

目的主流深度学习的目标检测技术对自然影像的识别精度依赖于锚框设置的好坏,并使用平行于坐标轴的正框表示物体位置,而遥感影像中地物目标具有尺寸多变、分布密集、长宽比悬殊且朝向不定的特点,更宜通过与物体朝向一致的斜框表示其位置。本文试图结合无锚框和斜框检测技术,在遥感影像上实现高精度目标识别。方法使用斜框标注能够更为紧密地贴合目标边缘,有效减少识别干扰因素。本文基于单阶段无锚框目标检测算法:一阶全卷积目标检测网络(fully convolutional one-stage object detector, FCOS),通过引入滑动点结构,在遥感影像上实现高效率、高精度的斜框目标检测。与FCOS的不同之处在于,本文改进的检测算法增加了用于斜框检测的两个分支,通过在正框的两邻边上回归滑动顶点比率产生斜框,并预测斜框与正框的面积比以减少极端情况下的检测误差。结果在当前最大、最复杂的斜框遥感目标检测数据集DOTA(object detection in aerial images)上对本文方法进行评测,使用ResNet50作为骨干网络,平均精确率(mean average precision, mAP)达到74.84%,相比原始正框FCOS算法精度提升了33.02%,相比于YOLOv3(you only look once)效率提升了38.82%,比斜框检测算法R~3Det(refined rotation RetinaNet)精度提升了1.53%。结论实验结果说明改进的FCOS算法能够很好地适应高分辨率遥感倾斜目标识别场景。

关键词：深度学习遥感影像无锚框特征提取多尺度特征融合倾斜目标检测

基于改进D-LinkNet模型的高分遥感影像道路提取研究

在线全文

学校读者我要写书评

暂无评论

计算机工程 2021年第9期47卷 288-296页

作者：张立恒王浩薛博维何立明吕悦长安大学信息工程学院西安710064 西安中科星图空间数据技术有限公司西安710000

针对高分影像中的道路信息易受植被阴影、高楼建筑物、河流等非道路信息干扰的问题,提出一种高分遥感影像道路提取模型。在中心区域引入channel-spatial双注意力机制捕获道路信息的全局特征依赖,并基于原始模型DICE+BCE损失函数,构建新... 详细信息

针对高分影像中的道路信息易受植被阴影、高楼建筑物、河流等非道路信息干扰的问题,提出一种高分遥感影像道路提取模型。在中心区域引入channel-spatial双注意力机制捕获道路信息的全局特征依赖,并基于原始模型DICE+BCE损失函数,构建新型的超参数权重损失来优化网络模型中参数迭代的误差,改善道路分割的精度。按照1∶1、2∶1、3∶1、4∶1、5∶1这5种比值设定超参数权重比,通过调节超参数权值比获取模型最佳的道路分割性能。实验结果表明,与FCN-8s、U-Net等模型相比,改进D-LinkNet模型道路分割效果明显提升,能有效地规避因非道路因素对道路提取干扰而导致的“虚检”“漏检”“误检”的现象。

关键词：高分遥感影像双注意力机制全局特征依赖超参数权重损失道路分割

在线全文

学校读者我要写书评

暂无评论

10kV箱式变电站温度在线监测系统的研究

陕西电力 2016年第4期44卷 22-27页

作者：薛卜玮黄新波朱永灿西安工程大学电子信息学院陕西西安710048

箱式变电站室内温度及设备的温度变化都会影响其正常安全稳定运行。针对以往箱式变电站温度监测系统温度监测点不足、容易漏报以及传统测温方式存在的弊端,提出一套基于ZigBee无线测温和普通有线测温方式相结合的温度监测方案。在断路... 详细信息

箱式变电站室内温度及设备的温度变化都会影响其正常安全稳定运行。针对以往箱式变电站温度监测系统温度监测点不足、容易漏报以及传统测温方式存在的弊端,提出一套基于ZigBee无线测温和普通有线测温方式相结合的温度监测方案。在断路器触头、隔离开关旋转触头、母排压接头处等空间狭小、布线困难等常见热源部位安装ZigBee无线测温装置模块,在高压室、变压室、低压室等利于布线的地方选用价格更为低廉的有线测温装置。所设计的箱式变电站温度监测系统不但有效地解决了传统箱式变压器温度监测装置所存在的问题,实现了箱变温度全方位实时在线监测功能,同时还兼有低成本、低功耗、高可靠性等优点。

关键词：箱式变电站温度无线网络 ZigBee 在线监测

Secondary phase precipitate-induced localized corrosion of pure aluminum anode for aluminum–air battery

在线全文

学校读者我要写书评

暂无评论

International Journal of Minerals,Metallurgy and Materials 2023年第5期30卷 977-987页

作者： bowei Zhang Hezu Wang Yan Su Wenguang Yang xuelong Hao Zequn Zhang Fengqin Wang Wei xue Junsheng Wu Institute for Advanced Materials and Technology University of Science and Technology BeijingBeijing 100083China Southwest Institute of Technology and Engineering Chongqing 400039China National Center of Analysis and Testing for Nonferrous Metals and Electronic Materials Guobiao(Beijing)Testing&Certification Co.Ltd.Beijing 101407China

Understanding the influence of purities on the electrochemical performance of pure aluminum(Al)in alkaline media for Al–air batteries is ***,we comprehensively investigate secondary phase precipitate(SPP)-induced localized corrosion of pure Al in NaOH solution mainly based on quasi-in-situ and cross-section observations under scanning electron microscopy coupled with finite element *** experimental results indicate that Al–Fe SPPs appear as clusters and are coherent with the Al *** alkaline media,Al–Fe SPPs exhibit more positive potentials than the substrate,thus aggravating localized galvanic corrosion as cathodic ***,finite element simulation indicates that the irregular geometry coupled with potential difference produces the non-uniform current density distribution inside the SPP cluster,and the current density on the Al substrate gradually decreases with distance.

关键词： aluminum secondary phase precipitates cluster corrosion alkaline

Reverse-blocking modular multilevel converter for battery energy storage systems

在线全文

学校读者我要写书评

暂无评论

Journal of Modern Power Systems and Clean Energy 2017年第4期5卷 652-662页

作者： Xiaofeng YANG Yao xue bowei CHEN Yajie MU Zhiqin LIN Trillion Q.ZHENG Seiki IGARASHI School of Electrical Engineering Beijing Jiaotong UniversityBeijing 100044China Fuji Electric Co. Ltd.Tokyo 141-0032Japan

Energy storage systems with multilevel converters play an important role in modern electric power systems with large-scale renewable energy *** paper proposes a reverse-blocking modular multilevel converter for a battery energy storage system(RB-MMCBESS). Besides integrating distributed low-voltage batteries to medium or high voltage grids, with the inherited advantages of traditional MMCs, the RB-MMC-BESS also provides improved DC fault handling capability. This paper analyzes such a new converter configuration and its operating principles. Control algorithms are developed for AC side power control and the balancing of battery state of charge. The blocking mechanism to manage a DC pole-topole fault analyzed in depth. Comprehensive simulation results validate both the feasibility of the RB-MMC-BESS topology and the effectiveness of the control and fault handling strategies.

关键词： Reverse blocking Modular multilevel converter Battery energy storage system SOC control Fault blocking

Novel Modular Multilevel Converter Against DC Faults for HVDC Applications

在线全文

学校读者我要写书评

暂无评论

CSEE Journal of Power and Energy Systems 2017年第2期3卷 140-149页

作者： Xiaofeng Yang Yao xue bowei Chen Zhiqin Lin Trillion Q.Zheng Yan Li School of Electrical Engineering Beijing Jiaotong UniversityBeijing 100044China State Key Laboratory of Operation and Control of Renewable Energy&Storage Systems China Electric Power Research InstituteBeijing 100192China

In view of the DC fault current isolation deficiency for the conventional half-bridge sub-module(HBSM)based modular multilevel converter(MMC),this paper presents an improved MMC *** quasi reverse blocking submodules(QRBSMs)and current limit modules(CLMs)are employed to improve the DC fault handling capability for HVDC *** paper analyzes such a new converter configuration and operation *** the DC pole-to-pole short circuit fault is taken into consideration for further study,as well as the fault current blocking mechanism and quantitative relationship between system electrical stress and key *** validate the feasibility of the proposed topology and fault protection theory,extensive simulation results are *** is concluded that the QRB-MMC can effectively block the fault current under DC fault *** addition,CLMs play an important role in further accelerating fault current ***,QRB-MMC employs the original control and modulation strategies under normal operation conditions;thus,it further reduces the complexity of industry design.

关键词： Current limit module DC faults HVDC transmission modular multilevel converter quasi reverse blocking sub-module