Abstract:
The purposes of this research are to find and develop multimodal biometric system using hand geometry and iris pattern. This research composes of 3 biometric systems: a biometric system using hand geometry, a biometric system using iris pattern and a multimodal biometric system using hand geometry and iris pattern. The biometric system using hand geometry is a system that uses the physiological or shape of a hand to identify the identity of a person. The 21 features used in this research are the lengths of 5 fingers, the widths of 5 fingers at the 1/3, 1/2 and 2/3 of the length, and the width of a palm. S[subscript 1] distance function is used to match the features. The system tested with 96 users possesses 91.06% accuracy. The biometric system using iris pattern uses an iris pattern to identify the identity of a person. In this research, only the left and the right side of an iris pattern are used. They are divided into 24x24 block size with 6 overlapping pixels. Standard deviation (SD) of each block is used as a feature of the iris pattern, and hence resulting in 120 features. Absolute distance function is used in the matching process. When tested with 96 users, the system accuracy is 91.55%. The multimodal biometric system using hand geometry and iris pattern in this research is done by integrating the two systems in the matching score level. In each system, the person with least matching score is selected. The matching scores of hand geometry and iris pattern are normalized with the possible min-max scores of each system. The person who has less combined score is the identity of the user. The system is tested with 96 users and the result is 98.33% accuracy