Hull vector-based incremental learning of hyperspectral remote sensing images

被引:7
作者
Huang, Fenghua [1 ,2 ]
Yan, Luming [3 ]
机构
[1] Fuzhou Univ, Postdoctoral Programme Elect Sci & Technol, Fuzhou 350116, Peoples R China
[2] Yango Coll, Fuzhou 350015, Peoples R China
[3] Fujian Normal Univ, Coll Geog Sci, Fuzhou 350007, Peoples R China
来源
JOURNAL OF APPLIED REMOTE SENSING | 2015年 / 9卷
基金
中国国家自然科学基金;
关键词
hull vector; hyperspectral remote sensing images; binary detection theory-sequential minimal optimization classifier; class-incremental learning;
D O I
10.1117/1.JRS.9.096022
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
To overcome the inefficiency of incremental learning for hyperspectral remote sensing images, we propose a binary detection theory-sequential minimal optimization (BDT-SMO) nonclass-incremental learning algorithm based on hull vectors and Karush-Kuhn-Tucker conditions (called HK-BDT-SMO). This method can improve the accuracy and efficiency of BDT-SMO nonclass-incremental learning for fused hyperspectral images. But HK-BDT-SMO cannot effectively solve class-incremental learning problems (an increase in the number of classes in the newly added sample sets). Therefore, an improved version of HK-BDT-SMO based on hypersphere support vector machine (called HSP-BDT-SMO) is proposed. HSP-BDT-SMO can substantially improve the accuracy, scalability, and stability of HK-BDT-SMO class-incremental learning. Ultimately, HK-BDT-SMO and HSP-BDT-SMO are applied to the classification of land uses with fused hyperspectral images, and the classification results are compared with other incremental learning algorithms to verify their performance. In nonclass-incremental learning, the accuracy of HSP-BDT-SMO and HK-BDT-SMO is approximately the same and is higher than the others, and the former has the best learning speed; while in class-incremental learning, HSP-BDT-SMO has a better accuracy and more continuous stability than the others and the second highest learning speed next to HK-BDT-SMO. Therefore, HK-BDT-SMO and HSP-BDT-SMO are excellent algorithms which are respectively suitable to nonclass and class-incremental learning for fused hyperspectral images. (C) 2015 Society of Photo-Optical Instrumentation Engineers (SPIE)
引用
收藏
页数:15
相关论文
共 28 条
[1]  
[Anonymous], 2013, THESIS
[2]  
Bai D., 2007, AERONAUT COMPUT TECH, V37, P23
[3]  
Cao Jian, 2014, Fire Control and Command Control, V39, P139
[4]   Spectral-Spatial Classification of Hyperspectral Image Based on Kernel Extreme Learning Machine [J].
Chen, Chen ;
Li, Wei ;
Su, Hongjun ;
Liu, Kui .
REMOTE SENSING, 2014, 6 (06) :5795-5814
[5]  
[冯佳 FENG Jia], 2009, [微电子学与计算机, Microelectronics & Computer], V26, P216
[6]  
Fenghua Huang, 2014, Journal of Multimedia, V9, P456, DOI 10.4304/jmm.9.3.456-462
[7]  
Guo K., 2008, COMMUN SCI TECHNOL H, V12, P169
[8]  
Hong Jiguang, 1984, Acta Automatica Sinica, V10, P22
[9]  
Huang F., 2014, SCI WORLD J, V2014, P13
[10]  
[黄文嘉 Huang Wenjia], 2012, [遥感信息, Remote Sensing Information], V27, P4