Design and implementation of an Autoencoder for the suppression of noise of different nature in color images
DOI:
https://doi.org/10.46842/ipn.cien.v27n1a04Keywords:
autoencoder, convolutions, neural networks, denoisingAbstract
This article focuses on the proposal of a neural network called Autoencoder for the suppression of various kinds of noise present in color images. There are several kinds of algorithms for denoising images, such as Convolutional Neural Networks (CNN), which require a large amount of data for their training and a greater computational complexity, although the biggest problem they present is that they commonly focus on a single kind of noise, sometimes causing that the image to not be processed properly and at the end of its treatment it contains corrupted pixels which cause the loss of important details within the image. For this reason, this proposal provides evidence that the use of Autoencoders for the denoising of several kinds of noises is feasible, through this article subjective and objective results will be shown that will determine the feasibility of using this kind of network neural.
References
A. Limshuenchuey, R. Duangsoithong, M. Saejia, “Comparison of Image Denoising using Traditional Filter and Deep Learning Methods”, 17th International Conference on Electrical Engineering/Electronics, 2020.
V. Pnomaryov, A.J. Rosales-Silva, F. Gallegos-Funes, “Fuzzy directional filter to remove impulsive noise from colour images”, IEICE Trans. Fundamentals, pp. 570-572, 2010.
S. Schluter, A. Sheppard, Kendra Brown, Image processing of multi-phase images obtained via x-ray microtomography: a review, American Geophysical Union, 2014.
S. Kaur, “Noise types and various removal techniques”, International Journal of Advanced Research in Electronics and Communication Engineering, vol. 4, pp. 226-230, 2015.
R. C. Gonzalez, R. E. Woods, Digital Image Processing, 4a ed., Pearson, 2018.
T. Remez, R. Giryes, A. M. Bronstein, “Class-Aware Fully Convolutional Gaussian and Poisson Denoising”, IEEE Transactions on Image Processing, vol. 27, n.º 11, 2018.
S. Agarwal, A. Agarwal, M. Deshmukh, Denoising Images with Varying Noises Using Autoencoders, Springer Nature Singapore, 2020.
S. Yu, J. Príncipe, “Understanding autoencoders with information theoretic concepts”, Neural Networks, vol. 117, pp. 104-123, 2019.
A. Pawar, “Noise reduction in images using autoencoders”, Proceedings of the Third International Conference on Intelligent Sustainable Systems, 2020.
J. Dawani, Hands-On Mathematics for Deep Learning, Packt, 2020.
T. Ye, T. Wang, K. McGuinness, Learning Multiple Views with Orthogonal Denoising Autoencoders, Springer International, pp. 313-324, 2016.
Best Artworks of all Time, Kaggle: Your Machine Learning and Data Science Community, https://www.kaggle.com/datasets/ikarus777/best-artworks-of-all-time (acc. dec. 2022).
B. Kumar-Boyatm, “A review paper: Noise models in digital image processing”, Signal and Image Processing: An International Journal (SIPIJ), vol. 6, pp. 63-75, 2015.
A. Patil, A. Pramod, K. Singh, “An Approach to Image Denoising Using Autoencoders and Spatial Filters for Gaussian Noise”, 11th International Conference on Cloud Computing, Data Science & Engineering, pp. 454-458, 2021
K. Zhang, W. Ren, W. Luo, “Deep Image Deblurring: A Survey”, International Journal of Computer Vision, 2022.
K. Zhang, W. Zuo, Y. Chen, “Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising”, IEEE Transactions on Image Processing, vol. 26, no. 7, 2017.
C. Liangyu, C. Xiaojie, Z. Xiangyu, “Simple Baselines for Image Restoration”, Computer Vision and Pattern Recognition, 2022.
S. W. Zamir, A. Arora, S. Khan, “Restormer: Efficient Transformer for High-Resolution Image Restoration”, Computer Vision and Pattern Recognition, 2021.
C. Wickramasinghe, D. Marino, M. Manic, ResNet Autoencoders for Unsupervised Feature Learning From High-Dimensional Data: Deep Models Resistant to Performance Degradation, IEEE, vol. 9, 2021.
D. Wang, W. Gan, C. Yan, Inception Model of Convolutional Auto-encoder for Image Denoising, Springer Nature Switzerland, pp. 174-186, 2020.
Downloads
Published
Issue
Section
License
Copyright (c) 2023 Instituto Politecnico Nacional
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
