Implementation of multimodal biometrics recognition system combined palm print and palm geometry features

The study of biometric for recognition system has been extensively evolved because this system has many advantages compared to conventional system. Palm geometry is one example of biometric feature that able to used for determine/recognize people. It say that each people have a unique feature of palm geometry that able to distinguish between one person which another. In our research, we are using image-processing technique to make a system that receive palm image of a person and determine identity of those person. We combine palm-print feature and palm geometry feature to create a multimodal biometric recognition system. Palm geometry feature is compute base on 16-point marker that already determine. While palm print compute using line detection in ROI of palm. Data sets used in this research are 300 palm images from 50 users. 6 images are taken from each user, 3 for enrolment (reference) image and the 3 remaining for testing image. We achieve system best performance by combine palm-print and palm-geometry features using 50:50 scoring fusion mechanism, with False Rejection Rate 2.67%, False Acceptance Rate 0.71% and Genuine Acceptance Rate 97.37% with threshold 92.64.