测绘学报 ›› 2018, Vol. 47 ›› Issue (9): 1238-1249.doi: 10.11947/j.AGCS.2018.20170476

• 摄影测量学与遥感 • 上一篇    下一篇

融合空谱特征和集成超限学习机的高光谱图像分类

谷雨1, 徐英2, 郭宝峰1   

  1. 1. 杭州电子科技大学通信信息传输与融合技术国防重点学科实验室, 浙江 杭州 310018;
    2. 杭州电子科技大学生命信息与仪器工程学院, 浙江 杭州 310018
  • 收稿日期:2017-08-24 修回日期:2018-05-08 出版日期:2018-09-20 发布日期:2018-09-26
  • 作者简介:谷雨(1982-),男,博士,副教授,研究方向为遥感图像处理,多源信息融合。E-mail:guyu@hdu.edu.cn
  • 基金资助:
    国家自然科学基金(61771177;61375011)

Hyperspectral Image Classification by Combination of Spatial-spectral Features and Ensemble Extreme Learning Machines

GU Yu1, XU Ying2, GUO Baofeng1   

  1. 1. Fundamental Science on Communication Information Transmission and Fusion Technology Laboratory, Hangzhou Dianzi University, Hangzhou 310018, China;
    2. College of Life Information Science and Instrument Engineering, Hangzhou Dianzi University, Hangzhou 310018, China
  • Received:2017-08-24 Revised:2018-05-08 Online:2018-09-20 Published:2018-09-26
  • Supported by:
    The National Natural Science Foundation of China (Nos. 61771177;61375011)

摘要: 为提高高光谱图像的分类精度,提出了一种融合空谱特征和集成超限学习机的高光谱图像分类方法。首先结合每个像素邻域的光谱信息提取空谱特征向量;考虑到高光谱相邻波段信息具有一定的相关性,先对提取的特征向量进行平均分组,然后从每个区间随机选择若干个波段进行组合,采用具有快速学习能力的超限学习机训练分类器。为提高分类模型的泛化能力,基于集成学习思想,对提取的空谱特征进行多次抽样,训练得到多个弱分类器,最后采用投票表决法得到用于高光谱图像分类的强分类器。采用3个典型高光谱数据进行了分类试验,试验结果表明,提出的算法总体分类精度较优,尤其当训练样本数较少时能取得较高的分类精度。提出的算法具有可调参数少、训练速度快、分类精度高等优点,具有广阔的应用前景。

关键词: 高光谱图像分类, 空谱特征, 超限学习机, 集成学习, 特征抽样

Abstract: To improve hyperspectral image classification accuracy,a classification method based on combination of spatial-spectral features and ensemble extreme learning machines is proposed in this paper.First,a spatial-spectral feature vector for each pixel is extracted using its neighboring information. Considering the strong correlation relationship between neighboring bands in a hyperspectral image,average grouping is performed for the extracted features,and a certain number of bands are first selected randomly from each interval and then combined to form a new feature with fewer dimensions.Extreme learning machine which can be trained fast is used to train a classifier.To improve the generalization performance of the learned model,several rounds of sampling are carried out based on ensemble learning theory,and several weak classifiers are trained and then combined to build a strong classifier using majority vote method.The classification experiments are performed using three typical hyperspectral image datasets,and the experimental results demonstrate that,the proposed algorithm can achieve preferable results compared with the state-of-the-art classifiers.It can achieve better classification accuracies when fewer training samples are used.The proposed algorithm has the advantages of few adjustable parameters,fast training speed,and high classification accuracy,and can be applied in many areas.

Key words: hyperspectral image classification, spatial-spectral feature, extreme learning machine, ensemble learning, feature sampling

中图分类号: