PERSON IDENTIFICATION USING BIOMETRIC MODALITIES

Abstract

Biometric modalities provides us the study of characteristics related to physiological traits for the verification or identification of a person. Unimodal biometric arises and disadvantages of security, accuracy, performance and data handing. Using multimodal modalities, security many shortcomings imperfection, storage space will be much higher but accuracy, performance perfection, storage space and rate data handing are still a subject of research and discussion. In our current research study we are working with multimodal biometric modalities of face and thumb for the validation of any individual to obtain high security impression recognition with modalities offace and thumb impression. System will better accuracy using two be able to train face and thumb impression of persons, and for testing and validation both modalities are necessary to input or define. We are using Discrete Cosine Transform algorithm which collects features point of the sample, furthermore the data will be classified by the Support Vector Machine Technique to give the correct collected dataset. There outcome. Another contribution of our project is our own lack of datasets which can provide us both facial images as well as thumb were impression of the same person. The results have better accuracy rate and two modalities necessary to input for validation which (face and thumb impression) are makes the system more secure an applicable in many desired cases.