Indexed by:
Abstract:
For hyperspectral image (HSI) classification, it is very important to learn effective features for the discrimination purpose. Meanwhile, the ability to combine spectral and spatial information together in a deep level is also important for feature learning. In this letter, we propose an unsupervised feature learning method for HSI classification, which is based on recursive autoencoders (RAE) network. RAE utilizes the spatial and spectral information and produces high-level features from the original data. It learns features from the neighborhood of the investigated pixel to represent the whole local homogeneous area of the image. In addition, to obtain more accurate representation of the investigated pixel, a weighting scheme is adopted based on the neighboring pixels, where the weights are determined by the spectral similarity between the neighboring pixels and the investigated pixel. The effectiveness of our method is evaluated by the experiments on two hyperspectral data sets, and the results show that our proposed method has a better performance.
Keyword:
Reprint Author's Address:
Email:
Source :
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
ISSN: 1545-598X
Year: 2017
Issue: 11
Volume: 14
Page: 1928-1932
2 . 8 9 2
JCR@2017
3 . 9 6 6
JCR@2020
ESI Discipline: GEOSCIENCES;
ESI HC Threshold:118
JCR Journal Grade:2
CAS Journal Grade:3
Cited Count:
WoS CC Cited Count: 48
SCOPUS Cited Count: 73
ESI Highly Cited Papers on the List: 0 Unfold All
WanFang Cited Count:
Chinese Cited Count:
30 Days PV: 10
Affiliated Colleges: