文章摘要
吕飞,韩敏.基于深度极限学习机的高光谱遥感影像分类研究[J].,2018,58(2):166-173
基于深度极限学习机的高光谱遥感影像分类研究
Hyperspectral remote sensing image classification based on deep extreme learning machine
  
DOI:10.7511/dllgxb201802009
中文关键词: 高光谱遥感影像  深度学习  极限学习机  遥感影像分类
英文关键词: hyperspectral remote sensing image  deep learning  extreme learning machine  remote sensing image classification
基金项目:国家自然科学基金资助项目(61374154);国家自然科学基金委科学仪器基础研究专项资助项目(51327004)
作者单位
吕飞,韩敏  
摘要点击次数: 238
全文下载次数: 445
中文摘要:
      高光谱遥感数据越来越普及并为人们广泛使用,基于高光谱遥感数据的地面物体精确分类是高光谱遥感技术的核心应用之一.针对高光谱遥感影像的分类问题,提出一种基于深度极限学习机(D-ELM)的分类方法.该方法利用一种新的深度学习模型——深度极限学习机对高光谱遥感影像进行分类,并与基于极限学习机(ELM)、支持向量机(SVM)、核极限学习机(ELMK)分类方法进行了比较分析.研究结果表明:相对于ELM、SVM、ELMK分类方法,D-ELM分类方法能够更加准确地挖掘高光谱遥感影像的空间分布规律,提高分类的准确度.
英文摘要:
      Hyperspectral remote sensing data are more and more popular and widely used. The accurate classification of surface objects based on hyperspectral remote sensing data is one of the core applications of hyperspectral remote sensing technology. For the classification problem of hyperspectral remote sensing image, a classification method is proposed based on deep extreme learning machine (D-ELM). This method uses a new depth learning model, that is D-ELM to classify hyperspectral remote sensing images. It is compared with extreme learning machine(ELM), support vector machine (SVM), kernal extreme learning machine(ELMK) classification methods. The results show that D-ELM classification method can excavate the spatial distribution regularity of the hyperspectral remote sensing image more accurately, and improve the accuracy of classification with respect to ELM, SVM and ELMK classification methods.
查看全文   查看/发表评论  下载PDF阅读器
关闭