An intelligent deep learning LSTM-DM tool for finger vein recognition model USING DSAE classifier

In the current scenario, Biometric authentication plays a crucial role in the security process by relying on the inimitable genetic characteristics. The process is carried over by registering and comparing the data in the database. Being a vast field it addresses various looms like finger vein, palm vein, hand vein, finger print, palm print, etc. Deep Learning technique is applied for this type of approach based on the requirements. Thus a new intelligent Finger-Vein Recognition Model is proposed which adapts the Deep Learning (DL) techniques for Biometric Authentication Systems with a Decision-Making Tool (DMT). This proposed model enables recognition of Finger Vein related data by making use of a Long-Short-Term-Memory (LSTM-DMT) model adapting DL technique for the finger vein recognition system. Since classification of image is an essential process in this entire authentication system, a decision-making tool using a Deep Stacked Auto Encoder is used for the classification of different vein features that exist in big data. This entire model is designed in such a way that finger-vein Image Acquisition, preprocessing techniques, technique for extracting the features, matching of acquired images with database images, obtaining genuine images and lastly the evaluation of accuracy is carried out. A comprehensive experimental study of result is conceded guarantee betterment of proposed method. The final experimental upshot highlights the advanced better-quality results of proposed method in the matter of diverse measure's.