Gabor wavelet based blood vessel segmentation in retinal images using kernel classifiers

The retina is an important field in the medical treatment of pathologies. Segmentation of the vasculature in retinal fundus images plays an important role in the diagnosis of many eye diseases such as hypertension, arteriosclerosis and blindness caused by diabetes. Comparison of kernel classifiers for the vessel segmentation is presented in this work. Kernel based classifier such as Support Vector Machine and Relevance Vector Machine is used to segment the vessels by classifying each pixel as vessel or nonvessel, based on the pixel's feature vector. The feature vectors are composed of the pixel's intensity and the Gabor wavelet responses measured at different scales. The method's performance is evaluated on publicly available databases of color fundus images with reference to the ground truth image provided in the database. The performance of the Segmentation is analyzed in terms of Specificity, Sensitivity and Segmentation Accuracy.