English

应用科学学报 ›› 2011, Vol. 29 ›› Issue (6): 626-630.doi: 10.3969/j.issn.0255-8297.2011.06.012

• 信号与信息处理 • 上一篇    下一篇

图Laplacian半监督特征加权用于高光谱波段选择

黄睿， 陈玲   

  1. 上海大学通信与信息工程学院，上海200072
  • 收稿日期:2011-03-11 修回日期:2011-07-19 出版日期:2011-11-30 发布日期:2011-11-30
  • 通信作者: 黄睿，博士，讲师，研究方向：模式识别、图像处理、遥感信息智能处理，E-mail：huangr@shu.edu.cn
  • 作者简介:黄睿，博士，讲师，研究方向：模式识别、图像处理、遥感信息智能处理，E-mail：huangr@shu.edu.cn
  • 基金资助:

    国家自然科学基金(No.61001162)资助

Semi-supervised Feature Weighting Using Graph Laplacian for Hyperspectral Band Selection

HUANG Rui, CHEN Ling   

  1. School of Communication and Information Engineering, Shanghai University, Shanghai 200072, China
  • Received:2011-03-11 Revised:2011-07-19 Online:2011-11-30 Published:2011-11-30

PDF

1653

被引次数

4

摘要：

 提出一种利用图Laplacian实现半监督波段选择的方法. 该方法首先将标记样本类别信息引入图Laplacian，接着通过广义特征值求解确定投影变换矩阵，最后采用载荷因子对变换矩阵进行系数分析，对波段重要性赋以权值并排序. 实验比较了多种波段选择算法，结果表明算法能更好地利用标记样本的类别信息和大量非标记样本中的局部结构信息，性能优于多种波段选择方法.

关键词: 半监督特征加权, 图Laplacian, 波段选择, 高光谱数据分类

Abstract:

A semi-supervised feature weighting using graph Laplacian is proposed for hyperspectral band selection. The method first constructs the graph Laplacian modified by the label information. The projection matrix is obtained by solving a generalized eigen-problem. The corresponding matrix coefficients are analyzed using the loading factors to assign weights to the original bands. Experiments with hyperspectral data sets are carried out to make comparison among several band selection algorithms. The results show that the proposed method can achieve the best performance as it makes good use of class information from the labeled samples and local structure clues hidden in numerous unlabeled ones.

 

Key words: semi-supervised feature weighting, graph Laplacian, band selection, hyperspectral data classification

中图分类号: