<![CDATA[Retinal vascular pattern is very unique and has very good pattern differences between one individual and the others, so it can be argued that retinal image can be one of the best in the world of biometrics. This research create a system that can recognize retinal images using Gray Level Co-occurrence Matrix (GLCM) feature extraction technique and normalized Euclidean distance measurement techniques, which the image of the sample used was a normal retinal image of Messidor dataset. Based on testing of GLCM parameters (Angular Second Moment, Contrast, Entropy, and Inverse Difference Moment), distance, angle, and the number of images in the database, the largest accuracy of retinal image recognition is equal to 85% at the time of testing by using 45° angle, distance of 5 pixels, and an image in the database.]]>
Copyrights © 2016
