Noise reduction using spherical fuzzy metrics: Theory and applications
MEASUREMENT, cilt.292, 2026 (SCI-Expanded, Scopus)
- Yayın Türü: Makale / Tam Makale
- Cilt numarası: 292
- Basım Tarihi: 2026
- Doi Numarası: 10.1016/j.measurement.2026.123627
- Dergi Adı: MEASUREMENT
- Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Compendex, INSPEC, Academic Search Ultimate (EBSCO), Engineering Source (EBSCO), Materials Science & Engineering Collection (ProQuest), Technology Collection (ProQuest)
- Kocaeli Üniversitesi Adresli: Evet
Özet
In this study, the spherical fuzzy metric structure, obtained by extending spherical fuzzy set theory to fuzzy metric spaces, is applied to the problem of noise reduction in color images. In this context, a filtering method based on spherical fuzzy metrics is proposed, specifically for the removal of salt-and-pepper noise. In the proposed approach, the RGB components of color images are evaluated within the spherical fuzzy metric framework, with the aim of modeling the relationships between noisy pixels and their neighbors via fuzzy metrics. The performance of the proposed method is evaluated using various noise densities ranging from 1% to 50% on different images selected from the USC-SIPI and Kodak datasets. In comparative analyses, classical filters such as the Mean and Median are considered, along with different filtering approaches such as bilateral, Non-Local Means, and DnCNN. In the performance evaluation, MAE, PSNR, SSIM, and UQI are used to jointly examine the methods' behavior in terms of pixel-level error, structural similarity, and overall image quality. Additionally, comprehensive sensitivity analyses are conducted to determine the effect of the parameters used in spherical fuzzy metrics on the results. The results show that the proposed methods exhibit different performance characteristics under varying noise conditions, in particular, they can deliver competitive results for certain evaluation metrics at low noise densities. It is observed that as noise density increases, the performance of the methods varies depending on the metric used and the filtering approach being compared. These results demonstrate that spherical fuzzy metrics provide a useful framework for modeling uncertainty in the problem of noise reduction in color images and for developing an alternative filtering approach tailored to different noise conditions.