Features-Based Retinal Vessel Segmentation Methodologies: A Survey Sumathi T.*, Dr. Vivekanandan P.**, Dr. Balaji Ravikanth*** *Research Scholar, Department of Computer Science and Engineering, CEG, Anna University, Chennai, India **Professor of Eminence, Department of Chemical Engineering, Anna University, Chennai, India ***Senior Consultant, Radiology and Oncology Imaging, MRI and PET Divisions, Apollo Specialty Hospital, Chennai, India Online published on 23 March, 2017. Abstract The effective diagnosis and analysis of human retinal vessel provide information about the nature and severity of many retinalas well as the non-retinal diseases. The morphological attributes of retinal blood vessels are utilized for diagnosis and treatment of many diseases. Since the retinal vasculature is more complex, it is a tedious task to segment them manually and it has motivated the development of various automatic vessel segmentation methodologies. In this paper, a survey is made on recent retinal vasculature extraction techniques in perspective of features utilized for segmentation. Additionally, it also discusses about general classification of retinal vessel segmentation techniques. Most of the existing techniques use DRIVE database for performance evaluation since the database contains manually labeled segmentations. The performance of the algorithms discussed in this paper is compared and analyzed on DRIVE database by utilizing measures including accuracy, specificity and sensitivity. Top Keywords Retinal vessel, feature extraction, supervised classification, Retinal imaging, vessel segmentation, Retinal Databases. Top |