Evaluation of the segmented vascular structures of the retina of our eye obtained through fundus photography using an Unsupervised Learning algorithm. Two open-source databases of the retinal images (DRIVE and STARE) are used. K - Means Clustering Algorithm is used for the segmentation of the retinal images. MATLAB r2020b environment was employed for feature extraction and image segmentation. The accuracy of the segmentation is evaluated for both the database as well as the algorithms. A simple GUI was developed for convinience of evaluation.
DRIVE and STARE Dataset obtained from - https://www.medicmind.tech/retinal-image-databases#:~:text=Digital%20Retinal%20Images%20for%20Vessel%20Extraction%20%28DRIVE%29%20database,images%20in%20this%20database%20is%20565%20X%20584.
Add all files into the path of MATLAB before running
Run 'Project_retina.fig'