中文    English
当前位置: 本站首页 » 科学研究 » 科研成果 » 正文

Scene classification based on a hierarchical convolutional sparse auto-encoder for high spatial resolution imagery

发布日期:2017-03-17 16:07:11 阅读次数:[5638]次 作者:

核心提示:来源出版物: INTERNATIONAL JOURNAL OF REMOTE SENSING

作者: Han, XB (Han, Xiaobing); Zhong, YF (Zhong, Yanfei); Zhao, B (Zhao, Bei); Zhang, LP (Zhang, Liangpei)

来源出版物: INTERNATIONAL JOURNAL OF REMOTE SENSING  卷: 38  期: 2  页: 514-536  DOI: 10.1080/01431161.2016.1266059  出版年: 2017

摘要: Efficiently representing and recognizing the semantic classes of the subregions of large-scale high spatial resolution (HSR) remote-sensing images are challenging and critical problems. Most of the existing scene classification methods concentrate on the feature coding approach with handcrafted low-level features or the low-level unsupervised feature learning approaches, which essentially prevent them from better recognizing the semantic categories of the scene due to their limited mid-level feature representation ability. In this article, to overcome the inadequate mid-level representation, a patch-based spatial-spectral hierarchical convolutional sparse auto-encoder (HCSAE) algorithm, based on deep learning, is proposed for HSR remote-sensing imagery scene classification. The HCSAE framework uses an unsupervised hierarchical network based on a sparse auto-encoder (SAE) model. In contrast to the single-level SAE, the HCSAE framework utilizes the significant features from the single-level algorithm in a feedforward and full connection approach to the maximum extent, which adequately represents the scene semantics in the high level of the HCSAE. To ensure robust feature learning and extraction during the SAE feature extraction procedure, a 'dropout' strategy is also introduced. The experimental results using the UC Merced data set with 21 classes and a Google Earth data set with 12 classes demonstrate that the proposed HCSAE framework can provide better accuracy than the traditional scene classification methods and the single-level convolutional sparse auto-encoder (CSAE) algorithm.

版权所有:测绘遥感信息工程国家重点实验室   
联系地址: 中国·武汉市珞瑜路129号   邮编: 430079   E-mail:liesmars@whu.edu.cn
Tel/Fax:027-68778969(办公室) 027-68778229(国际交流办公室)027-68778525(研究生管理办公室)