Related Experiment Video
Updated: Apr 16, 2026
12:28
Quantification of Vascular Parameters in Whole Mount Retinas of Mice with Non-Proliferative and Proliferative Retinopathies
Published on: March 12, 2022
4.5K
Comparative study of retinal vessel segmentation based on global thresholding techniques
Temitope Mapayi1, Serestina Viriri1, Jules-Raymond Tapamo2
1School of Mathematics, Statistics & Computer Science, University of KwaZulu-Natal, Durban 4000, South Africa.
Computational and Mathematical Methods in Medicine
|March 21, 2015
Summary
Preprocessing retinal fundus images is crucial for accurate vessel segmentation. This study compares global thresholding with phase congruency and CLAHE, finding that careful selection of preprocessing, thresholding, and postprocessing methods is key for optimal retinal vessel segmentation.
Area of Science:
- Ophthalmology
- Medical Imaging
- Computer Vision
Background:
- Retinal fundus image acquisition can be affected by noise, uneven contrast, and illumination.
- Accurate retinal vessel segmentation is essential for diagnosing various eye conditions.
Purpose of the Study:
- To develop and compare different vessel segmentation techniques for retinal images.
- To evaluate the impact of preprocessing methods like phase congruency and contrast limited adaptive histogram equalization (CLAHE) on segmentation performance.
Main Methods:
- Global thresholding was employed for vessel segmentation.
- Phase congruency and CLAHE were utilized as preprocessing techniques.
- The performance of different combinations of preprocessing, global thresholding, and postprocessing was compared.
Main Results:
- The study demonstrated that preprocessing significantly impacts retinal vessel segmentation accuracy.
- Different combinations of preprocessing, global thresholding, and postprocessing yield varying segmentation results.
- No single combination universally outperformed others, highlighting the need for careful selection.
Conclusions:
- The choice of preprocessing, global thresholding, and postprocessing techniques must be carefully considered for effective retinal vessel segmentation.
- Optimizing these combined techniques is critical for achieving high-quality segmentation results in retinal fundus images.
