(Ministry of Education, General Directorate of Education in Babylon, Iraq)(2) Ali Abdulmunim Al-kharaz
(ITM department, Technical College of Management, Middle Technical University, Iraq)*corresponding author
AbstractImage denoising is a substantial section in the preprocessing stage, especially in medical images. This study proposed a hybrid denoising model for salt-and-pepper removal in grayscale images. The framework uses a U-Net convolutional neural network, modified to perform preliminary denoising, and the Alternating Direction Method (ADM) to refine the structure iteratively. A corrupted pixel location is first determined using an adaptive thresholding scheme. The model is trained with a composite loss function that combines pixel-wise reconstruction accuracy (MSE) and perceptual similarity, as measured by the Structural Similarity Index (SSIM). Tests conducted on benchmarks (e.g., Kodak24, Set14, DIV2K, and TID2013) show that the proposed method surpasses traditional filters and state-of-the-art deep learning models, e.g., FFDNet and DnCNN. The quantitative results are Peak Signal-to-Noise Ratio (PSNR) 32.45 dB, SSIM 0.92 against 30 percent salt-and-pepper noise, and the average speed of inference is 6.2 ms, showing improvements over baseline approaches in performance and appearance. The main innovation is combining a noise-aware adaptive detection step with a specially designed U-Net framework and ADM-sided refinement, achieving better edge preservation and robustness to noise at any level. The framework displays a high potential for use in medical imaging, document recovery, and real-time surveillance.
KeywordsSalt-besides-Pepper Noise; Grayscale Image Denoising; Alternating Direction Method (ADM); Real-Time Image Restoration; Adaptive Thresholding
|
DOIhttps://doi.org/10.26555/ijain.v11i4.2158 |
Article metricsAbstract views : 95 | PDF views : 7 |
Cite |
Full Text Download
|
References
[1] B. Smolka, D. Kusnik, M. Smolka, M. Kawulok, and B. Cyganek, “Color image denoising: a hybrid approach for mixed Gaussian and impulsive noise,” in Real-Time Image Processing and Deep Learning 2024, N. Kehtarnavaz and M. V. Shirvaikar, Eds., SPIE, Jun. 2024, p. 6. doi: 10.1117/12.3013424.
[2] B. Jiang, J. Li, Y. Lu, Q. Cai, H. Song, and G. Lu, “Eficient image denoising using deep learning: A brief survey,” Inf. Fusion, vol. 118, p. 103013, 2025, doi: 10.1016/j.inffus.2025.103013.
[3] R. S. Jebur, M. H. B. M. Zabil, D. A. Hammood, and L. K. Cheng, “A comprehensive review of image denoising in deep learning,” Multimed. Tools Appl., vol. 83, no. 20, pp. 58181–58199, 2024, doi: 10.1007/s11042-023-17468-2.
[4] E. A. H. Hernandez, Y. Cao, and N. Kehtarnavaz, “Deep learning architecture search for real-time image denoising,” in Proc.SPIE, May 2022, p. 1210205. doi: 10.1117/12.2620349.
[5] S. Guan et al., “Adaptive median filter salt and pepper noise suppression approach for common path coherent dispersion spectrometer,” Sci. Rep., vol. 14, no. 1, p. 17445, 2024, doi: 10.1038/s41598-024-66649-y.
[6] J. Gao, L. Li, X. Ren, Q. Chen, and Y. M. Abdul-Abbass, “An effective method for salt and pepper noise removal based on algebra and fuzzy logic function,” Multimed. Tools Appl., vol. 83, no. 4, pp. 9547–9576, 2024, doi: 10.1007/s11042-023-15469-9.
[7] S. N. Ali, S. B. Shuvo, M. I. S. Al-Manzo, A. Hasan, and T. Hasan, “An End-to-End Deep Learning Framework for Real-Time Denoising of Heart Sounds for Cardiac Disease Detection in Unseen Noise,” IEEE Access, vol. 11, pp. 87887–87901, 2023, doi: 10.1109/ACCESS.2023.3292551.
[8] S. Katta, P. Singh, D. Garg, and M. Diwakar, “A Hybrid Approach for CT Image Noise Reduction Combining Method Noise-CNN and Shearlet Transform,” Biomed. Pharmacol. J., vol. 17, no. 3, pp. 1875–1898, 2024, doi: 10.13005/bpj/2991.
[9] G. Al Shammari and N. Parveen, “A Dual-Channel Deep Learning Framework For Real-Time Detection Of Zero-Day Attacks Using Cnn-Lstm Hybrid Networks,” Nanotechnol. Perceptions, vol. 20, no. S13, pp. 1410–1440, 2024, doi: 10.62441/nano-ntp.v20iS13.92.
[10] R. Thakur and S. Raut, “Grayscale Image Colorization Using Deep Learning: A Case Study,” in Computing and Communications Engineering in Real-Time Application Development, Apple Academic Press, 2022, pp. 253–264. doi: 10.1201/9781003277217-20.
[11] S. N. Ali, S. B. Shuvo, and T. Hasan, “A robust deep learning framework for real-time denoising of heart sound,” TechRxiv, 2022, available at: Google Scholar.
[12] K. Radlak, L. Malinski, and B. Smolka, “Deep learning for impulsive noise removal in color digital images,” in Proc.SPIE, May 2019, p. 1099608. doi: 10.1117/12.2519483.
[13] D. M. Madhura and A. V Mohan, “Hybrid Image Denoising with Wavelet and Deep Learning Model,” J. Pharm. Negat. Results, vol. 13, no. 3, pp. 1282–1288, 2022, doi: 10.47750/pnr.2022.13.S03.201.
[14] S. Deshpande and S. B. Mukkanagoudar, “From Noise to Clarity: A Hybrid Approach for Image Denoising Using Traditional and Deep Learning Methods,” J. Comput. Sci. Technol. Stud., vol. 2, no. 2, pp. 39–52, 2020, doi: 10.32996/jcsts.2020.2.2.5.
[15] H. Zhu and M. K. Ng, “Structured dictionary learning for image denoising under mixed gaussian and impulse noise,” IEEE Trans. Image Process., vol. 29, pp. 6680–6693, 2020, doi: 10.1109/TIP.2020.2992895.
[16] S. Saponara, “Radar real-time image processing for machine perception,” in Proc.SPIE, May 2019, p. 109960N. doi: 10.1117/12.2519545.
[17] G. Halford, A. C. D. II, and C. P. Bailey, “A computationally efficient deep learning model for real-time image dehazing on edge devices,” in Proc.SPIE, May 2025, p. 134580C. doi: 10.1117/12.3053948.
[18] S. Adib, V. Vinogradov, and P. D. Gosling, “A Hybrid Digital Twin Framework for Real-Time Structural Damage Identification Using Physics-Based Models and Deep Learning,” 2025, doi: 10.2139/ssrn.5240191.
[19] Y. Gao, “Smart IoT with the hybrid evolutionary method and image processing for tumor detection,” Sci. Rep., vol. 15, no. 1, p. 31156, 2025, doi: 10.1038/s41598-025-16042-0.
[20] D. Mújica-Vargas, J. de Jesús Rubio, J. M. V. Kinani, and F. J. Gallegos-Funes, “An efficient nonlinear approach for removing fixed-value impulse noise from grayscale images,” J. Real-Time Image Process., vol. 14, no. 3, pp. 617–633, 2018, doi: 10.1007/s11554-017-0746-8.
[21] R. T. Cai et al., “A Denoising Method for Salt and Pepper Noise in Remote Sensing Based on Swin-Transformer Convolution U-Net and Filtering—FSCU-Net,” Earth Sp. Sci., vol. 12, no. 8, p. e2025EA004225, Aug. 2025, doi: 10.1029/2025EA004225.
[22] G. B. Sagara, G. B. Krishna, and S. S. Rawat, “Hybrid Deep Learning Framework for Real-Time Source Code Vulnerability Detection,” Commun. Appl. Nonlinear Anal., vol. 32, no. 7s, pp. 889–900, 2025, doi: 10.52783/cana.v32.3493.
[23] S. Lee, D. Kim, and S. Kim, “A deep learning-based template matching through other field of view infrared image pair for real-time mixed reality,” in Proc.SPIE, Jun. 2024, p. 1303403. doi: 10.1117/12.3013413.
[24] N. Cao and Y. Liu, “High-noise grayscale image denoising using an improved median filter for the adaptive selection of a threshold,” Appl. Sci., vol. 14, no. 2, p. 635, 2024, doi: 10.3390/app14020635.
[25] Q. Song and C. Gong, “Image reconstruction method for incomplete CT projection based on self-guided image filtering,” Med. Biol. Eng. Comput., vol. 62, no. 7, pp. 2101–2116, 2024, doi: 10.1007/s11517-024-03044-9.
[26] L. Alam and N. Kehtarnavaz, “Real-time generation of realistic defective wafer maps via deep learning network of CycleGAN,” in Real-Time Image Processing and Deep Learning 2023, SPIE, 2023, pp. 84–93. doi: 10.1117/12.2663364.
[27] S. Fu, X. Meng, F. Shen, H. Chen, and Y. Cao, “A Hybrid Deep Learning Framework for Real-Time Fault Diagnosis and Prediction of Elevator Systems,” Comput. Fraud Secur., vol. 2025, no. 2, pp. 265–275, 2025, doi: 10.52710/cfs.497.
[28] K. K. A. Viswanathan, “Evaluation of Deep Learning Architectures for Image Denoising,” Int. J. Sci. Res., vol. 14, no. 3, pp. 1521–1525, 2025, doi: 10.21275/SR25324143238.
[29] M. Arhami, A. Desiani, S. Yahdin, A. I. Putri, R. Primartha, and H. Husaini, “Contrast enhancement for improved blood vessels retinal segmentation using top-hat transformation and otsu thresholding.,” Int. J. Adv. Intell. Informatics, vol. 8, no. 2, pp. 210–223, 2022, doi: 10.26555/ijain.v8i2.779.
[30] K. Firdausy, O. Wahyunggoro, H. A. Nugroho, and M. B. Sasongko, “A new approach for sensitivity improvement of retinal blood vessel segmentation in high-resolution fundus images based on phase stretch transform.,” Int. J. Adv. Intell. Informatics, vol. 8, no. 3, pp. 299–312, 2022, doi: 10.26555/ijain.v8i3.914.
[31] A. A. Wirabudi, N. R. Fachrurrozi, P. Dorand, and M. Royhan, “Enhancement of images compression using channel attention and post-filtering based on deep autoencoder.,” Int. J. Adv. Intell. Informatics, vol. 10, no. 3, pp. 425–440, 2024, doi: 10.26555/ijain.v10i3.1499.
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
___________________________________________________________
International Journal of Advances in Intelligent Informatics
ISSN 2442-6571 (print) | 2548-3161 (online)
Organized by UAD and ASCEE Computer Society
Published by Universitas Ahmad Dahlan
W: http://ijain.org
E: info@ijain.org (paper handling issues)
andri.pranolo.id@ieee.org (publication issues)
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0