Advanced search
1 file | 821.65 KB Add to list

Training neural networks on artificially generated data: a novel approach to SAR speckle removal

Frieke Vancoillie (UGent) , Hans Lievens (UGent) , Isabelle Joos (UGent) , Aleksandra Pizurica (UGent) , Lieven Verbeke (UGent) , Robert De Wulf (UGent) and Niko Verhoest (UGent)
Author
Organization
Abstract
A neural network-based method for speckle removal in synthetic aperture radar (SAR) images is introduced. The method rests on the idea that a neural network learning machine, trained on artificially generated input-target couples, can be used to efficiently process real SAR data. The explicit plus-point of the method is that it is trained with artificially generated data, reducing the demands put on real input data such as data quality, availability and cost price. The artificial data can be generated in such a way that they fit the particular characteristics of the images to be denoised, yielding case-specific, high-performing despeckling filters. Acomparative study with three classical denoising techniques (Enhanced Frost (EF), Enhanced Lee (EL) and GammaMAP (GM)) and a wavelet filter demonstrated a superior speckle removal performance of the proposed method in terms of quantitative performance measures. Moreover, qualitative evaluation of the despeckled results was in favour of the proposed method, confirming its speckle removal efficiency.
Keywords
APERTURE RADAR IMAGES, BAYESIAN WAVELET SHRINKAGE, MULTIPLICATIVE NOISE, REDUCTION, STATISTICS

Downloads

  • (...).pdf
    • full text
    • |
    • UGent only
    • |
    • PDF
    • |
    • 821.65 KB

Citation

Please use this url to cite or link to this publication:

MLA
Vancoillie, Frieke et al. “Training Neural Networks on Artificially Generated Data: a Novel Approach to SAR Speckle Removal.” INTERNATIONAL JOURNAL OF REMOTE SENSING 32.12 (2011): 3405–3425. Print.
APA
Vancoillie, F., Lievens, H., Joos, I., Pizurica, A., Verbeke, L., De Wulf, R., & Verhoest, N. (2011). Training neural networks on artificially generated data: a novel approach to SAR speckle removal. INTERNATIONAL JOURNAL OF REMOTE SENSING, 32(12), 3405–3425.
Chicago author-date
Vancoillie, Frieke, Hans Lievens, Isabelle Joos, Aleksandra Pizurica, Lieven Verbeke, Robert De Wulf, and Niko Verhoest. 2011. “Training Neural Networks on Artificially Generated Data: a Novel Approach to SAR Speckle Removal.” International Journal of Remote Sensing 32 (12): 3405–3425.
Chicago author-date (all authors)
Vancoillie, Frieke, Hans Lievens, Isabelle Joos, Aleksandra Pizurica, Lieven Verbeke, Robert De Wulf, and Niko Verhoest. 2011. “Training Neural Networks on Artificially Generated Data: a Novel Approach to SAR Speckle Removal.” International Journal of Remote Sensing 32 (12): 3405–3425.
Vancouver
1.
Vancoillie F, Lievens H, Joos I, Pizurica A, Verbeke L, De Wulf R, et al. Training neural networks on artificially generated data: a novel approach to SAR speckle removal. INTERNATIONAL JOURNAL OF REMOTE SENSING. 2011;32(12):3405–25.
IEEE
[1]
F. Vancoillie et al., “Training neural networks on artificially generated data: a novel approach to SAR speckle removal,” INTERNATIONAL JOURNAL OF REMOTE SENSING, vol. 32, no. 12, pp. 3405–3425, 2011.
@article{853967,
  abstract     = {A neural network-based method for speckle removal in synthetic aperture radar (SAR) images is introduced. The method rests on the idea that a neural network learning machine, trained on artificially generated input-target couples, can be used to efficiently process real SAR data. The explicit plus-point of the method is that it is trained with artificially generated data, reducing the demands put on real input data such as data quality, availability and cost price. The artificial data can be generated in such a way that they fit the particular characteristics of the images to be denoised, yielding case-specific, high-performing despeckling filters. Acomparative study with three classical denoising techniques (Enhanced Frost (EF), Enhanced Lee (EL) and GammaMAP (GM)) and a wavelet filter demonstrated a superior speckle removal performance of the proposed method in terms of quantitative performance measures. Moreover, qualitative evaluation of the despeckled results was in favour of the proposed method, confirming its speckle removal efficiency.},
  author       = {Vancoillie, Frieke and Lievens, Hans and Joos, Isabelle and Pizurica, Aleksandra and Verbeke, Lieven and De Wulf, Robert and Verhoest, Niko},
  issn         = {0143-1161},
  journal      = {INTERNATIONAL JOURNAL OF REMOTE SENSING},
  keywords     = {APERTURE RADAR IMAGES,BAYESIAN WAVELET SHRINKAGE,MULTIPLICATIVE NOISE,REDUCTION,STATISTICS},
  language     = {eng},
  number       = {12},
  pages        = {3405--3425},
  title        = {Training neural networks on artificially generated data: a novel approach to SAR speckle removal},
  url          = {http://dx.doi.org/10.1080/01431161003749436},
  volume       = {32},
  year         = {2011},
}

Altmetric
View in Altmetric
Web of Science
Times cited: