Abstract
Automatic target recognition (ATR) in synthetic aperture radar (SAR) images requires algorithms that are robust to the unique characteristics of a SAR image. As detection and classification can be time-consuming for even a SAR expert, neural networks (NNs) have emerged as a powerful solution to the problem. NNs learn by studying example images and can adapt to target and environmental variations in subsequent images. This work compares the accuracy of a Feed Forward Neural Network (FFNN) and a Convolutional Neural Network (CNN) on SAR images. While both networks achieved good performance, the CNN outperformed the FFNN.
| Original language | English (US) |
|---|---|
| Title of host publication | EUSAR 2018 - 12th European Conference on Synthetic Aperture Radar, Proceedings |
| Publisher | Institute of Electrical and Electronics Engineers Inc. |
| Pages | 1022-1027 |
| Number of pages | 6 |
| Volume | 2018-June |
| ISBN (Electronic) | 9783800746361 |
| State | Published - Jan 1 2018 |
| Externally published | Yes |
| Event | 12th European Conference on Synthetic Aperture Radar, EUSAR 2018 - Aachen, Germany Duration: Jun 4 2018 → Jun 7 2018 |
Other
| Other | 12th European Conference on Synthetic Aperture Radar, EUSAR 2018 |
|---|---|
| Country | Germany |
| City | Aachen |
| Period | 6/4/18 → 6/7/18 |
Fingerprint
ASJC Scopus subject areas
- Signal Processing
- Instrumentation
Cite this
Comparison of feed forward neural networks and convolutional neural networks for SAR automatic target recognition. / Hunter, Heather; Graber, Hans C.
EUSAR 2018 - 12th European Conference on Synthetic Aperture Radar, Proceedings. Vol. 2018-June Institute of Electrical and Electronics Engineers Inc., 2018. p. 1022-1027.Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - Comparison of feed forward neural networks and convolutional neural networks for SAR automatic target recognition
AU - Hunter, Heather
AU - Graber, Hans C
PY - 2018/1/1
Y1 - 2018/1/1
N2 - Automatic target recognition (ATR) in synthetic aperture radar (SAR) images requires algorithms that are robust to the unique characteristics of a SAR image. As detection and classification can be time-consuming for even a SAR expert, neural networks (NNs) have emerged as a powerful solution to the problem. NNs learn by studying example images and can adapt to target and environmental variations in subsequent images. This work compares the accuracy of a Feed Forward Neural Network (FFNN) and a Convolutional Neural Network (CNN) on SAR images. While both networks achieved good performance, the CNN outperformed the FFNN.
AB - Automatic target recognition (ATR) in synthetic aperture radar (SAR) images requires algorithms that are robust to the unique characteristics of a SAR image. As detection and classification can be time-consuming for even a SAR expert, neural networks (NNs) have emerged as a powerful solution to the problem. NNs learn by studying example images and can adapt to target and environmental variations in subsequent images. This work compares the accuracy of a Feed Forward Neural Network (FFNN) and a Convolutional Neural Network (CNN) on SAR images. While both networks achieved good performance, the CNN outperformed the FFNN.
UR - http://www.scopus.com/inward/record.url?scp=85050503336&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85050503336&partnerID=8YFLogxK
M3 - Conference contribution
VL - 2018-June
SP - 1022
EP - 1027
BT - EUSAR 2018 - 12th European Conference on Synthetic Aperture Radar, Proceedings
PB - Institute of Electrical and Electronics Engineers Inc.
ER -