Retinal Health Investigation by Segmentation of Major and Minor Blood Vessels in Fundus Images for Diabetes Patients

Abstract

Introduction: The Major and Minor blood vessel and the optic disc are the important factors for finding the presence of retinal disease like diabetic retinopathy. Aim: To segment the blood vessel using Major vessel extraction and Minor vessel classification methods. Materials and Methods: The retinal blood vessels are analysed from the images, for determining the changes in the blood vessels depending on the vessel branching pattern, width and density. High pass filter is used for easily segmenting the major vessels and also to make the retinal nerves lighter and the eye background dark to avoid fovea. A morphological operation like the top hat is performed for zooming and shrinking the nerves to adjust the outer structure of the eye. The Minor vessels are obtained by using the derivative of Gaussian filter without any crack or without any breaks in the structure of the retinal nerves which is used to determine the presence of blood clot in the eye. Gabor Filter is used to strengthening the pixels of the image. Conclusion: The segmented outputs of the Major vessel and Minor vessels are together added to get the resultant image of the eyes. Thus extracted and classified vessels can be sent to an intelligent artificial neural network for comparison with threshold values for diagnosing the problems due to high glucose level, cardiac disorders, etc.

Keywords

• diabetic retinopathy
• extraction and classification
• morphological operation prediction of eye health
• retinal disease