Blood Vessel Segmentation Using FCM–STSA Method for Retinal Fundus Images

Retinal blood vessel segmentation can be defined as separating blood vessels from retinal fundus images for detecting various kinds of ophthalmic ailments. This paper presents an effective blood vessel segmentation approach using the Sine Tree–Seed Algorithm (STSA) and fuzzy C-mean (FCM) clustering. The key objective of the proposed approach is to enhance the efficacy of the segmentation phase so that blood vessels are extracted efficiently from retinal fundus images. To enhance the quality of images, the adaptive histogram equalization (AHE) technique is applied to raw dataset images to enhance their contrast. Moreover, the average filtration technique is used for denoising images and makes them more suitable for segmentation. Finally, STSA and FCM clustering approaches segregate blood veins in retinal images. The efficacy of the suggested FCM–STSA-based blood vessel segmentation approach is analyzed on DRIVE and STARE datasets in MATLAB software. The simulated outcomes prove the proposed approach's supremacy over existing methods.