Design of single and hybrid classifier Based score level fusion for Multimodal biometric recognition

Abstract

Biometrics refers to the usage of physiological or behavioural newlinecharacteristics for identifying individuals With increasing progress in newlinetechnology security plays a vital role Biometric recognition is a reliable newlinetechnique to validate the identification of individuals based on physiological newlineor behavioural traits In recent times multimodal biometrics where two or newlinemore modalities are used for increasing recognition has been the main focus newlinefor several scholars. The decision of correct feature is an important mission in newlinerecognizing individuals Three biometric modalities are chosen viz newlinefingerprint face and iris in the present research newlineThe objectives of this research work are to extract the features from newlinemultiple biometric traits and generate the scores for each trait To develop a newlinescore level fusion technique for multimodal biometric recognition using newlineArtificial Bee Colony ABC based Neural Network NN for improving the newlinerecognition accuracy and reducing the False Acceptance Rate FAR and newlineFalse Rejection Rate FRR To design a multi classifier based score level newlinefusion for multimodal biometrics recognition using hybridization technique newlineand compare the accuracy values newline newline