Hyperspectral image classification based on average spectral-spatial features and improved hierarchical-ELM

•A new average spectral-spatial feature.•We proposed a fast hyperspectral image classification method.•The classification results are determined by a spatial decision. The major purpose of this study is to investigate a fast and accurate classification method for hyperspectral images (HSI). First, t...

Full description

Saved in:
Bibliographic Details
Published in:Infrared physics & technology 2019-11, Vol.102, p.103013, Article 103013
Main Authors: Le, Ba Tuan, Ha, Thai Thuy Lam
Format: Article
Language:English
Subjects:
Average spectral- spatial
Classification
Hyperspectral images (HSI)
Improved hierarchical-ELM
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:•A new average spectral-spatial feature.•We proposed a fast hyperspectral image classification method.•The classification results are determined by a spatial decision. The major purpose of this study is to investigate a fast and accurate classification method for hyperspectral images (HSI). First, the average spectral images are calculated based on spectral-spatial information. Second, the objects mentioned above are classified by using the improved hierarchical extreme learning machine. Finally, the final classification results are determined by spatial decisions. The proposed method is better than other advanced classification methods in training time and accuracy and yields satisfactory results.
ISSN:1350-4495
1879-0275
DOI:10.1016/j.infrared.2019.103013