Half iris Gabor based iris recognition

摘要

Many researchers realised that iris as biometrics is the most accurate means for recognition since iris of human contain uniquepatterns. Hence in this study, we used Gabor filter as feature extraction to evaluate the possibility of extracting half iris for recognition with Support Vector Machine as classifier. Initial findings based on experimental analysis conducted showed that the proposed method is suitable with half iris as input based on performance attained specifically FAR and FRR of 0.21% and ERR of 99.56%. Thus, this technique could be used to overcome occlusion due to eyelids and eyelashes during segmentation.