A METHOD AND SYSTEM FOR GENERATING AND CLASSIFYING A HYBRID FEATURE VECTOR FOR BIOMETRIC RECOGNITION

The present disclosure relates to a method and system for generating and classifying a Hybrid Feature Vector for Biometric Recognition. More particularly, in the present disclosure, a technique is provided where a fusion of Kekre's Median Codebook Generation (KMCG), Kekre's Fast Codebook Generation (KFCG) and CNN-Based Feature vectors is carried out to increase the accuracy with reduced memory time and learning time. In the present disclosure, feature vectors are first extracted through a Mechanism such as Vector Quantization, Transform Domain analysis, and Wavelet Analysis, which are then combined with CNN based feature vector, and finally the resultant feature vectors are used for training artificial neural network. This way, the accuracy of CNN's custom filters is increased for feature extraction resulting in improved overall accuracy of classification.