Local Binary Pattern based Multimodal Biometric Recognition using Ear and FKP with Feature Level Fusion

Biometrics is a mode of popularity and identification affirmation that uses physical attributes of a particular person that are impossible or as a minimum hard to mask. Finger-Knuckle-Print and ear print are examined to be maximum reliable physiological bio metric methods. These days a Finger Knuckle Print authentication based totally bio metric system has been efficaciously applied in various sectors. These unique Finger Knuckle Print and ear print authentication offers in reveals its user image that matches with its very own template that is stored inside the database for verification. An ear print is a replica of the parts outside the ear which have impressed a particular surface [3]. Many variants of local binary patterns are widely used for feature extraction process due to their satisfactory performance, for recognition purpose. An effective method of extracting feature patterns for Finger Knuckle Print (FKP) and ear has been proposed in this project. After extracting features of ear and FKP separately, they are fused by feature level fusion process which is done registration phase. The fused results of FKP and ear have been compared with the stored fused values in authentication phase. If the level of matching is achieved, that particular person gets authenticated to access the things else the person will be impostor [4]. This system gives more accuracy in less time. Challenges: The accuracy of authentication is improved in less duration. The chances of intruding is reduced.