Comparison of Training Strategies for Autoencoder-Based Monochromatic Image Denoising
Monochromatic images are used mainly in cases where the intensity of the received signal is examined. The identification of the observed objects as well as the estimation of intensity emitted by them depends largely on the precision of light measurement in image pixels. Unfortunately, this type of i...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-06-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/23/12/5538 |
| _version_ | 1827735680318439424 |
|---|---|
| author | Piotr Jóźwik-Wabik Krzysztof Bernacki Adam Popowicz |
| author_facet | Piotr Jóźwik-Wabik Krzysztof Bernacki Adam Popowicz |
| author_sort | Piotr Jóźwik-Wabik |
| collection | DOAJ |
| description | Monochromatic images are used mainly in cases where the intensity of the received signal is examined. The identification of the observed objects as well as the estimation of intensity emitted by them depends largely on the precision of light measurement in image pixels. Unfortunately, this type of imaging is often affected by noise, which significantly degrades the quality of the results. In order to reduce it, numerous deterministic algorithms are used, with Non-Local-Means and Block-Matching-3D being the most widespread and treated as the reference point of the current state-of-the-art. Our article focuses on the utilization of machine learning (ML) for the denoising of monochromatic images in multiple data availability scenarios, including those with no access to noise-free data. For this purpose, a simple autoencoder architecture was chosen and checked for various training approaches on two large and widely used image datasets: MNIST and CIFAR-10. The results show that the method of training as well as architecture and the similarity of images within the image dataset significantly affect the ML-based denoising. However, even without access to any clear data, the performance of such algorithms is frequently well above the current state-of-the-art; therefore, they should be considered for monochromatic image denoising. |
| first_indexed | 2024-03-11T01:57:11Z |
| format | Article |
| id | doaj.art-8a81696945a44aca80489936e0b72abb |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-11T01:57:11Z |
| publishDate | 2023-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-8a81696945a44aca80489936e0b72abb2023-11-18T12:32:31ZengMDPI AGSensors1424-82202023-06-012312553810.3390/s23125538Comparison of Training Strategies for Autoencoder-Based Monochromatic Image DenoisingPiotr Jóźwik-Wabik0Krzysztof Bernacki1Adam Popowicz2Faculty of Automatic Control, Electronics and Computer Science, Silesian University of Technology, Akademicka 16, 44-100 Gliwice, PolandFaculty of Automatic Control, Electronics and Computer Science, Silesian University of Technology, Akademicka 16, 44-100 Gliwice, PolandFaculty of Automatic Control, Electronics and Computer Science, Silesian University of Technology, Akademicka 16, 44-100 Gliwice, PolandMonochromatic images are used mainly in cases where the intensity of the received signal is examined. The identification of the observed objects as well as the estimation of intensity emitted by them depends largely on the precision of light measurement in image pixels. Unfortunately, this type of imaging is often affected by noise, which significantly degrades the quality of the results. In order to reduce it, numerous deterministic algorithms are used, with Non-Local-Means and Block-Matching-3D being the most widespread and treated as the reference point of the current state-of-the-art. Our article focuses on the utilization of machine learning (ML) for the denoising of monochromatic images in multiple data availability scenarios, including those with no access to noise-free data. For this purpose, a simple autoencoder architecture was chosen and checked for various training approaches on two large and widely used image datasets: MNIST and CIFAR-10. The results show that the method of training as well as architecture and the similarity of images within the image dataset significantly affect the ML-based denoising. However, even without access to any clear data, the performance of such algorithms is frequently well above the current state-of-the-art; therefore, they should be considered for monochromatic image denoising.https://www.mdpi.com/1424-8220/23/12/5538image denoisingGaussian noiseautoencoder |
| spellingShingle | Piotr Jóźwik-Wabik Krzysztof Bernacki Adam Popowicz Comparison of Training Strategies for Autoencoder-Based Monochromatic Image Denoising Sensors image denoising Gaussian noise autoencoder |
| title | Comparison of Training Strategies for Autoencoder-Based Monochromatic Image Denoising |
| title_full | Comparison of Training Strategies for Autoencoder-Based Monochromatic Image Denoising |
| title_fullStr | Comparison of Training Strategies for Autoencoder-Based Monochromatic Image Denoising |
| title_full_unstemmed | Comparison of Training Strategies for Autoencoder-Based Monochromatic Image Denoising |
| title_short | Comparison of Training Strategies for Autoencoder-Based Monochromatic Image Denoising |
| title_sort | comparison of training strategies for autoencoder based monochromatic image denoising |
| topic | image denoising Gaussian noise autoencoder |
| url | https://www.mdpi.com/1424-8220/23/12/5538 |
| work_keys_str_mv | AT piotrjozwikwabik comparisonoftrainingstrategiesforautoencoderbasedmonochromaticimagedenoising AT krzysztofbernacki comparisonoftrainingstrategiesforautoencoderbasedmonochromaticimagedenoising AT adampopowicz comparisonoftrainingstrategiesforautoencoderbasedmonochromaticimagedenoising |