Hierarchical feature exttratction for object recogition in complex SAR image using modified convolutional auto-encoder

Automatic target recognition is a crucial task for SAR remote sensing. Unlike other methods, the unsupervised representation learning based on deep architecture can obtain robust high-level features directly from raw data. A drawback of most unsupervised representation learning methods in SAR ATR is...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:IEEE International Geoscience and Remote Sensing Symposium proceedings S. 854 - 857
Hauptverfasser: Tian, S.R., Wang, C., Zhang, H.
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 01.07.2017
Schlagworte:
automatic target recognition(ATR)
Deep learning
Feature extraction
Image recognition
Kernel
Modified convolutional denoising autoencoder (MCSAE)
Noise reduction
Radar polarimetry
Remote sensing
Representation learning
Stacking
Synthetic aperture radar
synthetic aperture radar(SAR)
Target recognition
ISSN:2153-7003
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Automatic target recognition is a crucial task for SAR remote sensing. Unlike other methods, the unsupervised representation learning based on deep architecture can obtain robust high-level features directly from raw data. A drawback of most unsupervised representation learning methods in SAR ATR is that they only deal with amplitude images. In addition, many methods utlize a single layer architecture to extract pixel-level/mid-level features which are probably sensitive to condition variation. In this paper, a feature extraction method based on modified stacked convolutional denoising auto-encoder (MSCDAE) for complex SAR images is proposed, where convolutional kernels of MSCDAE are learned by 1-D modified denoising auto-encoders. By stacking the convolutional layers and pooling layers, high-level respresntation of objects are learned. The features are subsequently sent to a trained SVM for object classification. Experimetnal results demonstrate that the proposed method can provide a significant improvement in the ATR performance.
ISSN:2153-7003
DOI:10.1109/IGARSS.2017.8127087