Abstract
Unimodal biometrics systems face a diversity of problems. Multimodal biometric systems are proposed to solve these problems such as background noise and distortion. It used to enhance accuracy of the Identification system. This paper introduces a bimodal biometric recognition system. In this system, two traits are combined such as plamprint and iris. An approach for feature extraction of each modality is proposed by using texture based descriptor the Scale Invariant Feature Transform (SIFT). Different fusion techniques were tested at different levels. In feature fusion level, many feature vectors are used to compose a single feature vector, these features will be concatenated then to reduce the dimensionality of the feature, particle swarm optimization (PSO) is used as a feature selection technique. The proposed algorithm will be applied to the IITD database. By miming the voting technique, the proposed system generates excellent recognition results with less selected features.