journal article·2026-07-23

Color Representation and Prediction Uncertainty in Unsupervised Retinal Vessel Segmentation

Patrycja Kwiek, Małgorzata Jakubowska

Applied Sciences· Volume 16· Issue 15
Publisher page

Abstract

This study investigates the influence of color representation, preprocessing, and thresholding strategies on unsupervised retinal vessel segmentation using computationally efficient classical image processing techniques. Nineteen image channels derived from RGB, HSV, HSL, YUV, YCrCb, CIELAB, and grayscale representations were evaluated in combination with five thresholding algorithms on the DRIVE dataset. The results demonstrate that luminance-dominant channels consistently outperform chrominance-based representations. The Moments (Tsai) thresholding algorithm provided the most robust performance across image representations, while contrast enhancement using CLAHE and white top-hat filtering significantly improved segmentation quality. In addition, the weighted combination of RGB channels was characterized by a dominant green-channel contribution (wG ≈ 0.88–0.90) and a minimal blue-channel contribution (wB ≈ 0–0.02), reflecting the spectral properties of retinal vasculature. The study highlights the central role of luminance information and demonstrates that task-specific optimization of color representations improves both segmentation accuracy and reliability. Although deep learning methods currently achieve state-of-the-art performance in retinal vessel segmentation, the results show that substantial optimization remains possible using classical, interpretable, and computationally efficient approaches.
Source: crossref · Quality: baseline

Affiliations

Faculty of Materials Science and Ceramics, AGH University of Krakow, al. Mickiewicza 30, 30-059 Kraków, Poland

Funding

AGH University of Krakow · Minister of Science and Higher Education

References (36)

Show source-provided references
1. Analysis of Retinal Blood Vessel Segmentation Techniques: A Systematic Survey10.1007/s11042-022-13388-9 ↗
2. Recent Trends and Advances in Fundus Image Analysis: A Review10.1016/j.compbiomed.2022.106277 ↗
3. A Review of Retinal Vessel Segmentation for Fundus Image Analysis10.1016/j.engappai.2023.107454 ↗
4. Bansal, A., Kubíček, J., Bahrová, J., Varyšová, A., Augustýnek, M., Penhaker, M., and Timkovič, J. (2026). Retinal Vessel Segmentation: A Comprehensive Review from Classical Methods to Deep Learning Advances (1982–2025). Adv. Intell. Syst., e202501279.10.1002/aisy.202501279 ↗
5. Retinal blood vessel segmentation by means of scale-space analysis and region growing
6. A Survey of Deep Learning for Retinal Blood Vessel Segmentation Methods: Taxonomy, Trends, Challenges and Future Directions10.1109/ACCESS.2022.3163247 ↗
7. Deep Learning for Retinal Vessel Segmentation: A Systematic Review of Techniques and Applications10.1007/s11517-025-03324-y ↗
8. Spatially Enhanced U-Net with Non-Local Attention Gates for Automated Retinal Vasculature and Foveal Avascular Zone Segmentation in Optical Coherence Tomography Angiography Images10.1016/j.engappai.2026.113956 ↗
9. Dense-PMSFNet: DenseNet pyramidal multi-scale fusion network for retinal vasculature and FAZ segmentation in OCTA images10.1109/TLA.2025.10930374 ↗
10. An Ensemble Classification-Based Approach Applied to Retinal Blood Vessel Segmentation10.1109/TBME.2012.2205687 ↗
11. U-Net: Convolutional Networks for Biomedical Image Segmentation
12. A Threshold Selection Method from Gray-Level Histograms10.1109/TSMC.1979.4310076 ↗
13. A New Criterion for Automatic Multilevel Thresholding10.1109/83.366472 ↗
14. Picture Thresholding Using an Iterative Selection Method10.1109/TSMC.1978.4310039 ↗
15. Moment-Preserving Thresholding: A New Approach10.1016/0734-189X(85)90133-1 ↗
16. Heckbert, P.S. (1994). Contrast Limited Adaptive Histogram Equalization. Graphics Gems, Elsevier.
17. Contrast Enhancement for Improved Blood Vessels Retinal Segmentation Using Top-Hat Transformation and Otsu Thresholding
18. Retinal Blood Vessel Segmentation Based on Heuristic Image Analysis10.2298/CSIS180220014B ↗
19. Mean Global Based on Hysteresis Thresholding for Retinal Blood Vessel Segmentation Using Enhanced Homomorphic Filtering10.1007/s11042-022-13517-4 ↗
20. Ali, N.H., Hamzah, N.A.A., Saad, N.M., Mahfooz, R., and Abdullah, A.R. (2023). Segmentation of Diabetic Retinopathy Based on Retinal Fundus Images Using Thresholding Technique. Proceedings of the Multimedia University Engineering Conference (MECON 2022), Atlantis Press.10.2991/978-94-6463-082-4_17 ↗
21. Mishra, A., Pandey, M., and Singh, L. (2023). Multi-Level Otsu Thresholding and Morphological Operations for Fundus Image Segmentation for Diabetic Retinopathy Identification. Proceedings of the 2023 10th IEEE Uttar Pradesh Section International Conference on Electrical, Electronics and Computer Engineering, UPCON, IEEE.10.1109/UPCON59197.2023.10434899 ↗
22. Shapar, V., and Rani, J. (2024). Retinal Vessel Segmentation Using Morphological Operations. Applied Data Science and Smart Systems, CRC Press.10.1201/9781003471059-55 ↗
23. A High Accuracy Segmentation Method for Retinal Blood Vessel Detection Based on Hybrid Filters and an Adaptive Thresholding10.1007/s12652-022-03893-y ↗
24. Diabetic Retinopathy Detection via Exudates and Hemorrhages Segmentation Using Iterative NICK Thresholding, Watershed, and Chi2 Feature Ranking10.1038/s41598-025-90048-6 ↗
25. Diagnosis of Retinal Detachment via Blood Vessel Analysis Using Multi Threshold Image Binarization10.54105/ijpmh.C1064.05040525 ↗
26. Babu, K.M., Chowdappa, K.B., Mayee, M.K., Rao, A.S., and Mounika, R. (2025). Diabetic Retinopathy: An Exploration of Retinal Blood Vessel Segmentation Using Multilayered Thresholding. Ann. Data Sci.10.1007/s40745-025-00614-y ↗
27. Machine learning driven hybrid method for retinal vessel segmentation
28. Retina Vessel Segmentation via Improved Search and Rescue Algorithm and Bayesian Optimization10.1007/s11042-026-21505-1 ↗
29. van Ridge-Based Vessel Segmentation in Color Images of the Retina10.1109/TMI.2004.825627 ↗
30. Lakens, D. (2013). Calculating and Reporting Effect Sizes to Facilitate Cumulative Science: A Practical Primer for t-Tests and ANOVAs. Front. Psychol., 4.10.3389/fpsyg.2013.00863 ↗
31. The Bonferonni and Šidák Corrections for Multiple Comparisons
32. SciPy 1.0: Fundamental Algorithms for Scientific Computing in Python10.1038/s41592-019-0686-2 ↗
33. Array Programming with NumPy10.1038/s41586-020-2649-2 ↗
34. Matplotlib: A 2D Graphics Environment10.1109/MCSE.2007.55 ↗
35. Data Structures for Statistical Computing in Python
36. Locating Blood Vessels in Retinal Images by Piecewise Threshold Probing of a Matched Filter Response10.1109/42.845178 ↗