BFOA with varying population based feature selection and optimization in palm print authentication — A comparative study

This paper is about the comparative study of performances of different Bacterial Foraging Optimization Algorithms in feature selection and optimization in a new palm-print based biometric authentication system. The reference features are obtained from the Cross Wavelet Transformed (XWT) histograms of palm images with respect to reference palm image. A project cost is very low as it employs a scanner coupled to a PC, for acquiring the image of the human palm. A trained ANN is used to identify the subject, based on these features. Bacterial Foraging Optimization Algorithm(BFOA) and bacterial foraging optimization algorithm with varying population (named BFVPA) and their adaptive models are used to select the combination of features that results in the best performance in palm print based biometric identification systems. Extensive experimental investigations reveal that an authentication accuracy of more than 97.85% can be obtained with the proposed method.