Computer Engineering and Applications ›› 2021, Vol. 57 ›› Issue (10): 57-64.DOI: 10.3778/j.issn.1002-8331.2101-0427

Previous Articles     Next Articles

Application of Deep Learning in High Resolution Remote Sensing Image Scene Classification

ZENG Li, XU Huiying, CHEN Xiaohao, QIAN Xiaoliang   

  1. School of Electrical and Information Engineering, Zhengzhou University of Light Industry, Zhengzhou 450002, China
  • Online:2021-05-15 Published:2021-05-10

深度学习在高分遥感图像场景分类中的应用

曾黎,徐慧颖,陈晓昊,钱晓亮   

  1. 郑州轻工业大学 电气信息工程学院,郑州 450002

Abstract:

Scene classification of high resolution remote sensing image is committed to automatically identify the types of land use or cover, which has important application value in military and land resources exploration. High resolution remote sensing image scene classification method based on deep learning has achieved better results than traditional methods, and it is also a hot spot of current research. This paper summarizes and comprehensively evaluates these methods. Firstly, according to the different ways of supervision, the popular methods based on deep learning are analyzed. Secondly, the popular methods under different supervision methods are evaluated quantitatively on three open data sets. Finally, the characteristics of different supervision methods are summarized, and the future development trend is prospected.

Key words: high resolution remote sensing image, deep learning, scene classification, supervision method

摘要:

高分遥感图像场景分类致力于自动辨别土地利用或覆盖的类别,在军事和国土资源勘探等领域具有重要的应用价值。基于深度学习的高分遥感图像场景分类方法取得了比传统方法更好的效果,也是当前研究的热点,对此类方法进行归纳总结和综合评估。按照监督方式的不同,对基于深度学习的流行方法进行了逐类分析。对不同监督方式下的流行方法在三个公开数据集上进行了定量实验评估。总结了基于不同监督方式方法的特点,并对下一步发展趋势进行了展望。

关键词: 高分遥感图像, 深度学习, 场景分类, 监督方式