Object-oriented Classification of Polarimetric SAR Imagery based on Kernel Fisher Discriminant Dimensionality Reduction

Konferenz: EUSAR 2016 - 11th European Conference on Synthetic Aperture Radar
06.06.2016 - 09.06.2016 in Hamburg, Germany

Tagungsband: EUSAR 2016

Seiten: 4Sprache: EnglischTyp: PDF

Persönliche VDE-Mitglieder erhalten auf diesen Artikel 10% Rabatt

Autoren:
Cao, Han; Zhang, Hong; Wang, Chao; Liu, Meng; Wu, Fan (Key Laboratory of Digital Earth Science, Institute of Remote Sensing and Digital Earth, CAS, China)

Inhalt:
In this paper the Kernel Fisher Discriminant (KFD) analysis technique is introduced into polarimetric SAR (PolSAR) feature dimensionality reduction and object-oriented land cover classification. As a supervised nonlinear dimensionality reduction technique, KFD first maps the input data into a high dimensional feature space with a nonlinear mapping, and then linear discriminant analysis is performed in the space. Thus the latent nonlinear features useful for classification can be extracted indirectly. At last, classification is performed on these extracted features. The results show that by using the nonlinear features instead of original ones and linear feature extraction based methods, better classification performance can be achieved.