计算机与现代化 ›› 2010, Vol. 1 ›› Issue (3): 7-5.doi: 10.3969/j.issn.1006-2475.2010.03.003

• 图像处理 • 上一篇    下一篇

基于多通道分类合成的SAR图像分类研究

李长春1 ,2,冒亚明1,孙灏3,慎利1   

  1. 1.北京师范大学资源学院,北京 100875;2.河南理工大学测绘与国土信息工程学院,河南 焦作 454000;3.中国矿业大学测绘与空间信息工程研究所,江苏 徐州221008
  • 收稿日期:2009-12-15 修回日期:1900-01-01 出版日期:2010-03-20 发布日期:2010-03-20

LI Chang-chun1,2,MAO Ya-ming1,SUN Hao3,SHEN Li1   

  1. 1.College of Resource Science & Technology, Beijing Normal University, Beijing 100875, China; 2.School of Surveying & Land Information Engineering, Henan Polytechnic University, Jiaozuo 454000, China;3.Institute of Surveying and Spatial Information Engineering, China University of Mining and Technology, Xuzhou 221008, China
  • Received:2009-12-15 Revised:1900-01-01 Online:2010-03-20 Published:2010-03-20

摘要:

SAR 具有全天时、全天候工作能力,且能够提供高分辨率图像数据。SAR图像分类是SAR图像处理的关键步骤。目前,SAR图像分类多是基于单通道图像数据。多通道SAR数据极大地丰富了地物目标信息量,利用多通道数据进行分类,是SAR图像分类的重要发展方向。本文提出基于多通道分类合成的SAR图像分类算法。该算法首先利用SVM对不同通道的数据分别进行分类,然后利用粒度合成理论对不同的分类结果进行合并,最后实现多通道SAR数据图像分类。本文重点论述了利用该方法进行SAR图像分类的基本流程和步骤。最后,结合实验结果,证明了该算法的可行性和有效性。

关键词: 多通道, SAR, 图像分类, SVM, 粒度合成

Abstract:

SAR has the ability to work all weather, and can provide high resolution remote sensing images, so it is an important direction to study. The SAR image classification is a key step for SAR processing. At present, the study on SAR image classification is mostly based on singlechannel data, multichannel data has greatly enriched the information amount of ground target features, taking use of multichannel data to carry out SAR image classification is an important development direction. In the paper, the classification method of composing multichannel data classification results is proposed. Firstly, the method is to use different channel data to classify separately by SVM, then composes different classification results based on granularity composition theory. The paper mainly focuses on the basic flow and steps of this method. In the end, an experiment of using singlechannel data to classify and merging multichannel data features to classify is done, the result proves the method of this paper is viable.

Key words: multichannel data, SAR, image classification, SVM, quotient space granularity composition