SAR Images Despeckling Using Convolutional Neural Networks with Stochastic Distances

Pedro Henrique Salles Arouck de Souza; José Hiroki Saito; Francisco Fambrini

doi:10.22456/2175-2745.150936

Authors

Pedro Henrique Salles Arouck de Souza Centro Universitário Campo Limpo Paulista (UNIFACCAMP) https://orcid.org/0009-0009-1068-012X
José Hiroki Saito Centro Universitário Campo Limpo Paulista (UNIFACCAMP) https://orcid.org/0000-0003-0602-1980
Francisco Fambrini Universidade Virtual do Estado de São Paulo (UNIVESP) https://orcid.org/0000-0002-8841-5803

DOI:

https://doi.org/10.22456/2175-2745.150936

Keywords:

convolutional neural networks, noise, speckle, SAR images

Abstract

This work presents a study on speckle noise filtering in SAR images using convolutional neural networks. The objective is to develop an accessible and efficient solution to improve the quality of images degraded by this type of noise, facilitating their analysis and interpretation. The proposed methodology involves the using synthetic SAR-like data, the training of a convolutional neural network, and the evaluation of results using traditional metrics such as PSNR and SSIM. Additionally, stochastic distances were integrated into the loss function of the convolutional neural network, enabling a more detailed analysis of the preservation of the statistical properties of the filtered images. The results indicate that the model effectively reduces speckle noise while preserving the structural details of the image. Furthermore, comparisons with recent studies on the topic showed that the proposed solution achieves competitive performance.

Downloads

Download data is not yet available.

References

[1] ARSET. An Introduction to Synthetic Aperture Radar (SAR) and Its Applications. [S.l.], 2024. Applied Remote Sensing Training Program (ARSET). Disponível em: https://appliedsciences.nasa.gov/get-involved/training/english/arset-introduction-synthetic-aperture-radar-sar-and-its-applications.

[2] FRACASTORO, G. et al. Deep learning methods for synthetic aperture radar image despeckling: An overview of trends and perspectives. IEEE Geoscience and Remote Sensing Magazine, v. 9, n. 2, p. 29–51, 2021.

[3] DABHI, S. et al. Virtual SAR: A Synthetic Dataset for Deep Learning based Speckle Noise Reduction Algorithms. 2020.

[4] PASSAH, A. et al. SAR image despeckling using deep CNN. IET Image Processing, v. 15, n. 6, p. 1285–1297, 2021.

[5] PENNA, P. A. A.; MASCARENHAS, N. D. A. SAR speckle nonlocal filtering with statistical modeling of Haar wavelet coefficients and stochastic distances. IEEE Transactions on Geoscience and Remote Sensing, v. 57, n. 9, p. 7194–7208, 2019.

[6] EVANGELISTA, R. C. et al. A new Bayesian Poisson denoising algorithm based on nonlocal means and stochastic distances. Pattern Recognition, v. 122, p. 108363, 2022. ISSN 0031-3203.

[7] SANTOS, C. A. N. Redução de ruído speckle em imagens de ultrassom com filtragem não-local e distâncias estocásticas. Dissertação (Master’s thesis) — UFSCAR, 2017.

[8] CHENG, G.; HAN, J.; LU, X. Remote sensing image scene classification: Benchmark and state of the art. Proceedings of the IEEE, v. 105, n. 10, p. 1865–1883, 2017.

[9] BRIGATO, L.; IOCCHI, L. A close look at deep learning with small data. In: 2020 25th International Conference on Pattern Recognition (ICPR). [S.l.: s.n.], 2021. p. 2490–2497.

[10] KINGMA, D. P.; BA, J. Adam: A method for stochastic optimization. In: BENGIO, Y.; LECUN, Y. (Ed.). 3rd International Conference on Learning Representations, ICLR 2015, San Diego, CA, USA, May 7-9, 2015, Conference Track Proceedings. [S.l.: s.n.], 2015.

[11] CHENG, G.; HAN, J.; LU, X. Remote sensing image scene classification: Benchmark and state of the art. CoRR, abs/1703.00121, 2017.