Hyperspectral Image Classification via Multiple-Feature-Based Improved Sparse Representation

被引：9

作者：

Li Feiyan ^{[1
]}

Huo Hongtao ^{[1
]}

Li Jing ^{[1
]}

Bai Jie ^{[1
]}

机构：

[1] Peoples Publ Secur Univ China, Informat Technol & Cyber Secur Acad, Beijing 100038, Peoples R China

来源：

ACTA OPTICA SINICA | 2019年 / 39卷 / 05期

关键词：

remote sensing; hyperspectral image; sparse representation; feature extraction; Gabor filter; local binary pattern (LBP);

D O I：

10.3788/AOS201939.0528004

中图分类号：

O43 [光学];

学科分类号：

070207 ; 0803 ;

摘要：

A multiple-feature-based improved sparse representation (MFISR) method is proposed herein for the classification of hyperspectral images. The spectral feature, Gabor feature, and local binary pattern (ITT) feature arc extracted from the hyperspectral image; subsequently, the sparse coefficients arc solved and a 2-paradigm constraint is added. These obtained coefficients arc used to determine the final class label of each test pixel. The experimental results demonstrate that the proposed MSIFR method exhibits excellent results for the detection of small samples, and its classification performance is stable and good.

引用

页数：9