2026
Special Issue: Selected Proceedings of the 1st International Conference on Artificial Intelligence for Sustainability in the Developing World (AISDW2025)

Special Issue: Selected Proceedings of the 1st International Conference on Artificial Intelligence for Sustainability in the Developing World

Quantitative Evaluation of Adaptive Threshold-Based Denoising for AWGN-Corrupted Images Using Multiresolution Discrete and Stationary Wavelet Transforms

Published 2026-04-19

  • Firas Mahmood Mustafa
    • ,
  • Jalal Sami Issa

Firas Mahmood Mustafa
Cybersecurity Eng., Dept., Technical College of Engineering, Duhok Polytechnic University, Kurdistan Region of Iraq

Jalal Sami Issa
Information Technology Dept., Technical College of Duhok, Duhok Polytechnic University, Kurdistan Region of Iraq , Information Technology Dept., Technical College of Informatics, Akre University for Applied Sciences, Kurdistan Region of Iraq

Abstract

The integrity of digital images is often degraded by Additive White Gaussian Noise (AWGN), necessitating efficient denoising solutions. This study presents a comprehensive, quantitative benchmark of wavelet-based denoising by systematically evaluating the synergy between transform types and thresholding strategies. The main contributions of this work are threefold. First, it conducts a large-scale comparative analysis of 117 wavelet filters across both the Discrete Wavelet Transform (DWT) and the Stationary Wavelet Transform (SWT). Second, it integrates these transforms with both hard and adaptive SURE Shrink soft thresholding, evaluating their performance under two AWGN levels (sigma=15, sigma=25) on standard test images using Peak Signal-to-Noise Ratio (PSNR), while acknowledging the importance of perceptual metrics such as SSIM for future evaluation. Third, the results establish that SWT consistently yields higher PSNR than DWT, and SURE Shrink outperforms hard thresholding. Their combination achieved the best results, with average PSNR improvements of 15.26% and 13.23% over DWT for low and high noise, respectively. While the study is limited by its use of a single metric and grayscale images, the findings position SWT paired with SURE Shrink soft thresholding as a promising and competitive approach for image restoration in the presence of AWGN. Additionally, the study is primarily experimental, and further theoretical modeling is identified as future work.

Keywords

Image Denoising, AWGN, Wavelet Transform, DWT, SURE Shrink, PSNR

PDF

References

  1. C. Liu and H. Xu, “Recent progress in digital image restoration techniques: A review,” Remote Sensing of Environment, 2023, doi:10.1016/j.rse.2023.113450.
  2. L. Evans and X. Chen, “A review of digital image processing techniques and future prospects,” International Journal of Computer Science & Information Technology (IJCSIT), vol. 16, no. 2, pp. 45–60, 2024, doi:10.5121/ijcsit.2024.16204.
  3. C. Tian, M. Zheng, W. Zuo, and L. Zhang, “Overview of research on digital image denoising methods,” Sensors, vol. 25, no. 8, p. 2615, 2025, doi:10.3390/s25082615.
  4. C. Liu and L. Zhang, “A novel denoising algorithm based on wavelet and non-local moment mean filtering,” Electronics, vol. 12, no. 6, p. 1461, 2023, doi:10.3390/electronics12061461.
  5. Y. Zhou, S. Chen, and Y. Lin, “Evaluation of non-local means filter performance in medical image denoising,” Biomedical Signal Processing and Control, vol. 74, p. 103541, 2022, doi:10.1016/j.bspc.2021.103541.
  6. C. Tian, M. Zheng, W. Zuo, and L. Zhang, “Multi-stage image denoising with wavelet transform,” arXiv preprint arXiv:2209.12394, 2022.
  7. A. Kumar, H. Tomar, V. K. Mehla, R. Komaragiri, and M. Kumar, “Stationary wavelet transform based ECG and image denoising,” ISA Transactions, vol. 111, pp. 395–405, 2021, doi:10.1016/j.isatra.2021.07.012.
  8. K. Ravi and A. Srinivas, “Integrating low-rank methods and wavelet shrinkage for hyperspectral image denoising,” International Journal of Remote Sensing, vol. 44, no. 2, pp. 378–400, 2023, doi:10.1080/01431161.2023.2154563.
  9. H. Huang and P. L. Dragotti, “WINNet: Wavelet-inspired invertible network for image denoising,” IEEE Transactions on Image Processing, vol. 30, pp. 6070–6084, 2021, doi:10.1109/TIP.2021.3105403.
  10. T. Zhang, H. Liu, and Y. Wang, “A new hybrid image denoising algorithm using adaptive thresholding and low-rank decomposition,” Scientific Reports, vol. 15, no. 1, p. 8223, 2025, doi:10.1038/s41598-025-82283-3.
  11. J. Wang, G. Liu, and Y. Sun, “Advances in wavelet-based denoising for biomedical imaging,” Biomedical Engineering Letters, vol. 11, no. 3, pp. 321–337, 2021, doi:10.1007/s13534-021-00223-w.
  12. S. Ahmed and P. Kumar, “Evaluation of quality metrics in wavelet-based denoising,” Image and Vision Computing, vol. 129, p. 104664, 2023, doi:10.1016/j.imavis.2023.104664.
  13. M. Kaur and K. Kaur, “A review on image denoising techniques using wavelet transform,” Multimedia Tools and Applications, vol. 79, no. 21, pp. 15381–15414, 2020, doi:10.1007/s11042-019-08384-8.
  14. L. Wang and Z. Zhang, “An enhanced exposure fusion technique for multi-exposure image sets,” Journal of Visual Communication and Image Representation, vol. 84, p. 103567, 2023, doi:10.1016/j.jvcir.2022.103567.
  15. C. Liu and H. Xu, “Recent progress in digital image restoration techniques: A review,” Remote Sensing of Environment, 2023, doi:10.1016/j.rse.2023.113450.
  16. B. M. Ferzo and F. M. Mustafa, “Digital image denoising techniques in wavelet domain with another filter: A review,” Academic Journal of Nawroz University, vol. 9, no. 1, pp. 158–176, 2020, doi:10.25007/ajnu.v9n1a587.
  17. Y. Zhou, S. Chen, and Y. Lin, “Evaluation of non-local means filter performance in medical image denoising,” Biomedical Signal Processing and Control, vol. 74, p. 103541, 2022, doi:10.1016/j.bspc.2021.103541.
  18. F. M. Mustafa, “Image enhancement based on the histogram equalization and multiresolution discrete stationary wavelet transform,” Academic Journal of Nawroz University, vol. 11, no. 2, pp. 50–59, 2022, doi:10.25007/ajnu.v11n2a1323.
  19. P. Gómez and M. Ortega, “Statistical models for threshold optimization in multiscale transforms,” Signal Processing, vol. 187, p. 108137, 2021, doi:10.1016/j.sigpro.2021.108137.
  20. R. Narayan and M. Singh, “Comparative analysis of PSNR and SSIM for image quality,” Procedia Computer Science, vol. 199, pp. 734–740, 2022, doi:10.1016/j.procs.2022.01.092.
  21. H. Li and X. Zhou, “Adaptive wavelet shrinkage methods in signal denoising: A comparative study,” Journal of Signal Processing Systems, vol. 95, no. 1, pp. 45–59, 2023, doi:10.1007/s11265-023-01775-9.
  22. A. Singh and D. Rani, “Modern trends in wavelet domain image compression and denoising,” Signal & Image Processing: An International Journal, vol. 11, no. 2, pp. 21–35, 2020, doi:10.5121/sipij.2020.11202.
  23. B. M. Ferzo and F. M. Mustafa, “Image denoising in wavelet domain based on thresholding with applying Wiener filter,” in 2020 International Conference on Computer Science and Software Engineering (CSASE), 2020, doi:10.1109/CSASE48920.2020.9142091.
  24. F. M. Mustafa, “Image enhancement based on the histogram equalization and multiresolution discrete stationary wavelet transform,” Academic Journal of Nawroz University, vol. 11, no. 2, pp. 50–59, 2022, doi:10.25007/ajnu.v11n2a1323.
  25. F. M. Mustafa, H. S. Abdullah, and A. Elci, “Image enhancement in wavelet domain based on histogram equalization and median filter,” Journal of Engineering Research, 2021.
  26. F. M. Al-Fiky, A. Elci, and A. Ramadhan, “Image denoising by median filter in wavelet domain,” The International Journal of Multimedia & Its Applications (IJMA), vol. 9, no. 1, 2017, doi:10.5121/ijma.2017.9104.

Downloads

Download data is not yet available.

Similar Articles

21-29 of 29

You may also start an advanced similarity search for this article.