计算机工程与应用 ›› 2009, Vol. 45 ›› Issue (5): 193-195.DOI: 10.3778/j.issn.1002-8331.2009.05.057

• 工程与应用 • 上一篇    下一篇

网格环境下分布式SVM遥感图像分类器模型研究

曾联明1,2,3,吴湘滨1,刘 鹏3   

  1. 1.中南大学 地学与环境工程学院,长沙 410083
    2.佛山科学技术学院 信息中心,广东 佛山 528000
    3.解放军理工大学 网格技术研究中心,南京 210016
  • 收稿日期:2008-08-05 修回日期:2008-10-15 出版日期:2009-02-11 发布日期:2009-02-11
  • 通讯作者: 曾联明

Distribute image classification model of RS based on grid and SVM method

ZENG Lian-ming1,2,3,WU Xiang-bin1,LIU Peng3   

  1. 1.School of Geosciences and Environmental Engineering,Central South University,Changsha 410083,China
    2.Information and Educational Technology Center,Foshan University,Foshan,Guangdong 528000,China
    3.Grid Research Center,The PLA University of Science and Technology,Nanjing 210016,China
  • Received:2008-08-05 Revised:2008-10-15 Online:2009-02-11 Published:2009-02-11
  • Contact: ZENG Lian-ming

摘要: 为了充分利用网格技术分布式、高性能、协同共享的能力,设计了一种基于网格和支持向量机的分布式图像分类器模型,采用网格计算技术,统筹网络运算资源,结合支持向量机在有限样本统计分类中的优势,探索网格技术在图像分类中的应用。以对遥感图像目标物体的特征提取为例,实现基于分布式计算的图像分类过程,基于.net环境的实验结果表明,该模型提高了数据密集型图像分类速度和处理效率。

Abstract: Distributed image classification model based on grid and support vector machine is put forward in this paper to make full use of grid characteristic,such as distribution,high-performance and collaborative sharing capabilities etc.New model utilizes grid computing technology,plans as a whole of network computing resources,combines with statistic classification advantage of support vector machine in finite sample space,probes into application of grid computation technology in image classification.Taking characteristic extraction of interesting ground object from RS(Remote Sensing) images as the example,image classification process based on distribution computation is implemented.Experimental results based on .Net environment show that new model improves classification speed and efficiency of data-intensive images.