EUVIMEDEUVIMED entdecken
Crossref

WHUM: A Hybrid Wavelet Transform-Homomorphic Filtering-Unsharp Masking Method for Chest X-Ray Enhancement

Diana Diana, Ermatita Ermatita, Abdiansah -

International Journal of Advanced Computer Science and Applications · 2026 · 17 (10)

DOI: 10.14569/ijacsa.2026.0171043 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Advanced Computer Science and Applications, 2026. Verfasst von Diana Diana und 2 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The quality of chest X-ray (CXR) images can be greatly impacted by noise, uneven illumination, and low contrast, potentially obscuring anatomical details associated with identifying lung diseases. In this study, we propose Wavelet Homomorphic Unsharp Masking (WHUM), a sequential image-enhancement approach that uses a combination of Wavelet Transform (WT), Homomorphic Filtering (HF), and Unsharp Masking (UM). Specifically, we apply WT first to suppress high-frequency noise while maintaining the image structure, then HF to improve the illumination distribution and intensity properties, and finally UM to enhance the local details and structure boundaries. In quantitative evaluation, WT was found to achieve a high structural similarity (SSIM = 0.9866) and PSNR (43.9404 dB) in relation to the input image, which suggests that the stage only slightly modifies the input image. HF decreased the entropy, average gradient, and RMS contrast to 6.9444, 0.1378, and 0.1482, and the subsequent UM stage elevated them to 7.0406, 0.2098, and 0.1557 (SSIM = 0.9435, PSNR = 33.3206 dB). When the final WHUM image is compared to the preprocessed input, which had an entropy of 7.2489, an average gradient of 0.1683, and an RMS contrast of 0.1773, it becomes apparent that the WHUM output has a higher average gradient but also lower entropy and RMS contrast. Since both SSIM and PSNR were calculated by considering the input image, they evaluate the degree of fidelity, rather than how the enhancement is of good quality; while entropy along with gradient statistics serve to characterize, rather than prove the enhancement has genuinely increased structural details. WHUM was tested against histogram equalization, CLAHE, and gamma correction. Both techniques were applied independently and against variations of the processing stages, on a set of 1,500 held-out test images, drawn from a 70/15/15 split of 10,000 images in total. As another secondary verification, the effect the enhancement has on the ResNet50 classification was also quantified, indicating a ResNet50 accuracy and macro-F1 of 87.6% and 0.874 respectively for WHUM, compared to 86.4% and 0.861 for absence of enhancement, and 87.2% and 0.869 for CLAHE, and a visual examination of a single example of each class (Pneumonia, Tuberculosis, COVID-19, Normal) showed rib margins and vascular texture were markedly sharper after the WHUM.

Zitieren

APA

Diana, D., Ermatita, E., & -, A. (2026). WHUM: A Hybrid Wavelet Transform-Homomorphic Filtering-Unsharp Masking Method for Chest X-Ray Enhancement. International Journal of Advanced Computer Science and Applications, 17(10). https://doi.org/10.14569/ijacsa.2026.0171043

Vancouver

Diana D, Ermatita E, - A. WHUM: A Hybrid Wavelet Transform-Homomorphic Filtering-Unsharp Masking Method for Chest X-Ray Enhancement. International Journal of Advanced Computer Science and Applications. 2026;17(10). doi:10.14569/ijacsa.2026.0171043

BibTeX
@article{Diana2026,
  author = {Diana, Diana and Ermatita, Ermatita and -, Abdiansah},
  title = {WHUM: A Hybrid Wavelet Transform-Homomorphic Filtering-Unsharp Masking Method for Chest X-Ray Enhancement},
  journal = {International Journal of Advanced Computer Science and Applications},
  year = {2026},
  volume = {17},
  number = {10},
  doi = {10.14569/ijacsa.2026.0171043}
}
Kostenlos starten →