In the past years, researchers have shown more and more interests in synthetic aperture radar (SAR) automatic target recognition (ATR), and many methods have been proposed and studied for radar target recognition. Recently, deep learning methods, especially deep convolutional neural networks (CNN) has proven extremely competitive in image and speech recognition tasks. In this paper, a deep CNN model has been proposed for SAR automatic target recognition. The proposed deep model named SARnet, has two stage convolutional-pooling layers and two full-connected layers. Due to the demand of requirement of large scale of the data in deep learning, we proposed an augmentation method to get a large scale database for the training of CNN model, by which the CNN model can learn more useful features through the large scale database. Experimental results on the MSTAR database show the effectiveness of the proposed model and has achieved encouraging results with a correct recognition rate of 95.68%.

SAR automatic target recognition based on deep convolutional neural network / Y. Xu, K. Liu, Z. Ying, L. Shang, J. Liu, Y. Zhai, V. Piuri, F. Scotti (LECTURE NOTES IN COMPUTER SCIENCE). - In: Image and Graphics / [a cura di] Y. Zhao, X. Kong, D. Taubman. - [s.l] : Springer Verlag, 2017 Dec 30. - ISBN 9783319715971. - pp. 656-667 (( Intervento presentato al 9. convegno ICIG tenutosi a Shanghai nel 2017 [10.1007/978-3-319-71598-8_58].

SAR automatic target recognition based on deep convolutional neural network

V. Piuri;F. Scotti
2017

Abstract

In the past years, researchers have shown more and more interests in synthetic aperture radar (SAR) automatic target recognition (ATR), and many methods have been proposed and studied for radar target recognition. Recently, deep learning methods, especially deep convolutional neural networks (CNN) has proven extremely competitive in image and speech recognition tasks. In this paper, a deep CNN model has been proposed for SAR automatic target recognition. The proposed deep model named SARnet, has two stage convolutional-pooling layers and two full-connected layers. Due to the demand of requirement of large scale of the data in deep learning, we proposed an augmentation method to get a large scale database for the training of CNN model, by which the CNN model can learn more useful features through the large scale database. Experimental results on the MSTAR database show the effectiveness of the proposed model and has achieved encouraging results with a correct recognition rate of 95.68%.
Data augmentation; Deep convolutional neural networks (CNN); Deep learning; Synthetic aperture radar (SAR); Theoretical Computer Science; Computer Science (all)
Settore INF/01 - Informatica
Settore ING-INF/05 - Sistemi di Elaborazione delle Informazioni
30-dic-2017
Book Part (author)
File in questo prodotto:
File Dimensione Formato  
(2017) SAR Automatic Target Recognition Based on Deep Convolutional Neural Network.pdf

accesso riservato

Tipologia: Publisher's version/PDF
Dimensione 1.01 MB
Formato Adobe PDF
1.01 MB Adobe PDF   Visualizza/Apri   Richiedi una copia
(2017) SAR Automatic Target Recognition Based on Deep Convolutional Neural Network_postprint_ok.pdf

accesso aperto

Tipologia: Post-print, accepted manuscript ecc. (versione accettata dall'editore)
Dimensione 4.32 MB
Formato Adobe PDF
4.32 MB Adobe PDF Visualizza/Apri
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/2434/549011
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 8
  • ???jsp.display-item.citation.isi??? 3
social impact