Abstract—Retina based person identification is perceived as the most secure and accurate authentication means among biometric systems. The superiority of this method stems from the fact that the retina is unique to every individual and it would remain the same throughout a person’s life. In this paper, a novel method for person identification is presented using color retinal images. It consists of a parabolic model fitting, feature extraction and finally the matching process. A model based on parabolic fit is developed and used on the main arcade of retinal vasculature. The openness angle of the fitted parabola is calculated and named as Major Arcade angle (MAA). The parameters of the parabola and MAA are considered as features and then these undergo matching criteria for authentication. Performance of the proposed method is tested on three standard databases: DRIVE, HRF, Messidor, one local hospital database, and two authentication databases: RIDB, VARIA, and some pathological images.