TUTCRIS - Tampereen teknillinen yliopisto

TUTCRIS

Performance Comparison of Learned vs. Engineered Features for Polarimetric SAR Terrain Classification

Tutkimustuotosvertaisarvioitu

Yksityiskohdat

AlkuperäiskieliEnglanti
Otsikko2019 41st Photonics & Electromagnetics Research Symposium (PIERS)
DOI - pysyväislinkit
TilaHyväksytty/In press - 2019
OKM-julkaisutyyppiA4 Artikkeli konferenssijulkaisussa
TapahtumaPhotonics & Electromagnetics Research Symposium - Rome, Italia
Kesto: 17 kesäkuuta 201920 lokakuuta 2019
http://piers.org/piers2019Rome/

Conference

ConferencePhotonics & Electromagnetics Research Symposium
LyhennettäPIERS 2019
MaaItalia
KaupunkiRome
Ajanjakso17/06/1920/10/19
www-osoite

Tiivistelmä

In this work, we propose to use learned features for terrain classification of Polarimetric Synthetic Aperture Radar (PolSAR) images. In the proposed classification framework, the learned features are extracted from sliding window regions using Convolutional Neural Networks (CNNs), and then they are used for the classification with the linear Support Vector Machine (SVM) classifier. The classification performance of the proposed approach is compared with numerous target decomposition theorems (TDs) as the engineered features tested with two classifiers: Collective Network of Binary Classifiers (CNBCs) and SVMs. The experimental evaluations over two commonly used benchmark AIRSAR PolSAR images, San Francisco Bay and Flevoland at L-Band, reveal that the classification performance of the learned features with CNNs outperforms the performance of the engineered features as TDs even the dimension of learned features is the quarter of the engineered features.