计算机应用 ›› 2011, Vol. 31 ›› Issue (08): 2092-2096.DOI: 10.3724/SP.J.1087.2011.02092

• 人工智能 • 上一篇    下一篇

基于一类支持向量机的高光谱影像地物识别

陈伟1,余旭初1,张鹏强1,王智超2,王鹤3   

  1. 1. 信息工程大学 测绘学院,郑州450052
    2. 65015部队,辽宁 大连116023
    3. 北京望神州科技有限公司 销售部,北京100020
  • 收稿日期:2011-01-24 修回日期:2011-03-08 发布日期:2011-08-01 出版日期:2011-08-01
  • 通讯作者: 陈伟
  • 作者简介:陈伟(1983-),男,浙江杭州人,博士研究生,主要研究方向:模式识别、高光谱遥感;余旭初(1963-),男,湖北罗田人,教授,博士生导师,博士,主要研究方向:摄影测量与遥感、模式识别;张鹏强(1978-),男,甘肃镇原人,讲师,博士,主要研究方向:摄影测量与遥感、序列影像处理。

Object recognition based on one-class support vector machine in hyperspectral image

Wei CHEN1,Xu-chu YU1,Peng-qiang ZHANG1,Zhi-chao WANG2,He WANG3   

  1. 1. Institute of Surveying and Mapping, Information Engineering University, Zhengzhou Henan 450052, China
    2. Unit 65105, Dalian Liaoning 116023, China
    3. Sales Department, Digital LandView Technology Company Limited, Beijing 100020, China
  • Received:2011-01-24 Revised:2011-03-08 Online:2011-08-01 Published:2011-08-01
  • Contact: Wei CHEN

摘要: 高光谱遥感影像具有丰富的光谱信息,在地物识别方面具有明显的优势。一类支持向量机(OCSVM)不仅保留了支持向量机的原有优势,而且只需要待识别类型的训练样本。为此提出了算法,通过数学模型选择、核函数设计与参数的自适应调整将OCSVM原理融入到高光谱影像的地物识别算法中,提高了识别的精度,降低了对训练样本的要求。最后利用两幅高光谱影像进行了实验分析,实验结果证明了所提算法的有效性。

关键词: 高光谱影像, 一类支持向量机, 支持向量数据描述, 地物识别, 参数选择

Abstract: The hyperspectral remote sensing image is rich in spectrum information, so it has advantages in object recognition. One-Class Support Vector Machine (OCSVM) not only holds the advantages of support vector machines but also only needs the train samples of the recognized objects. The algorithm proposed in this paper selected mathematical model, designed kernel function, adjusted parameter adaptively, and added the theory of OCSVM into the object recognition algorithm for hyperspectral image which improved the precision of recognition and reduced the demand of train samples. Lastly, the experiments were conducted on two hyperspectral images, and the results prove the validity of the proposed method.

Key words: hyperspectral imagery, One-Class Support Vector Machine (OCSVM), Support Vector Data Description (SVDD), object recognition, parameter optimization

中图分类号: