Non-negative Matrix Factorization Features from Spectral Signatures of AVIRIS Images

被引:9
作者
Kaarna, Arto [1 ]
机构
[1] Lappeenranta Univ Technol, Dept Informat Technol, FIN-53851 Lappeenranta, Finland
来源
2006 IEEE INTERNATIONAL GEOSCIENCE AND REMOTE SENSING SYMPOSIUM, VOLS 1-8 | 2006年
关键词
D O I
10.1109/IGARSS.2006.145
中图分类号
P [天文学、地球科学];
学科分类号
07 ;
摘要
In this study we use non-negative matrix factorization (NMF) in deriving feature vectors from a set of spectral signatures. The purpose is to demonstrate the differences between the NMF and PCA feature vectors. The experiments show that NMF feature vectors are providing local features in spectral domain compared to the holistic features of PCA.
引用
收藏
页码:549 / 552
页数:4
相关论文
共 7 条
[1]  
HIISILA H, 2004, P 4 IEEE INT C DAT M
[2]  
Hoyer PO, 2004, J MACH LEARN RES, V5, P1457
[3]   THE SPECTRAL IMAGE-PROCESSING SYSTEM (SIPS) - INTERACTIVE VISUALIZATION AND ANALYSIS OF IMAGING SPECTROMETER DATA [J].
KRUSE, FA ;
LEFKOFF, AB ;
BOARDMAN, JW ;
HEIDEBRECHT, KB ;
SHAPIRO, AT ;
BARLOON, PJ ;
GOETZ, AFH .
REMOTE SENSING OF ENVIRONMENT, 1993, 44 (2-3) :145-163
[4]   Learning the parts of objects by non-negative matrix factorization [J].
Lee, DD ;
Seung, HS .
NATURE, 1999, 401 (6755) :788-791
[5]  
Ramanath R, 2003, P 32 APPL IM PATT RE, P33
[6]   Improving non-negative matrix factorizations through structured initialization [J].
Wild, S ;
Curry, J ;
Dougherty, A .
PATTERN RECOGNITION, 2004, 37 (11) :2217-2232
[7]  
Yuan ZJ, 2005, LECT NOTES COMPUT SC, V3540, P333