Laser & Optoelectronics Progress, Volume. 56, Issue 9, 091001(2019)
Improved Algorithm for Nonnegative Matrix Factorization and Endmember Extraction Based on Data Simplification
Figures & Tables(4)
Fig. 1. Flow chart of NMF mixed pixel decomposition algorithm improved by data simplification
Fig. 3. Endmember extraction results of MVC-NMF algorithm. (a) Before data simplification; (b) after data simplification
Table 1. Spectral angles between endmembers extracted by MVC-NMF algorithm and standard spectra in USGS spectral library before and after data simplificationrad
View tableTable 1. Spectral angles between endmembers extracted by MVC-NMF algorithm and standard spectra in USGS spectral library before and after data simplificationrad
Mineral specie Before data simplification After data simplification Alunite 0.0872 0.0866 Buddingtonite 0.1386 0.1384 Sphene 0.2106 0.2107 Chalcedony 0.1532 0.1542 Muscovite 0.0832 0.0835 Nortronite 0.2042 0.2052 Montmorillonite 0.1114 0.1117 Kaolinite#1 0.2358 0.2356 Desert varnish 0.3825 0.3842 Operation time/s 2686.3 539.8
Tools
Get Citation
Copy Citation Text
Jun Xu, Xuhong Wang, Cailing Wang. Improved Algorithm for Nonnegative Matrix Factorization and Endmember Extraction Based on Data Simplification[J]. Laser & Optoelectronics Progress, 2019, 56(9): 091001
Paper Information
Category: Image Processing
Received: Oct. 10, 2018
Accepted: Nov. 23, 2018
Published Online: Jul. 5, 2019
The Author Email: Jun Xu (3225393639@qq.com)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20