Identification of Finger Vein Images using Artificial Neural Network in comparison of Accuracy with K-means Algorithm
M.S. Sriram, I. Sudha · 2023
The goal of this recommended study is to identify and perform finger vein recognition of an individual to ensure the Digital security system in smart homes, industries and banks using two different machine learning (ML) techniques and compare the overall performance of selected classifiers. For this research, neural networks namely Artificial Neural Network and K-means classifiers are chosen. This phase involves selection of collection of data, training and testing of selected data with suggested classifiers as Artificial Neural Network and k-means over Pattern Recognition. For SPSS analysis, the outcome of two classifiers is categorized as two groups and each group consists of 20 samples with a G-power pre-test score of 80% and CI-95% is used. The selected Artificial Neural Network Algorithm shows improved digital security through finger vein recognition with the accuracy of 96.7100% and K-means classifier gained accuracy of 95.0410% The value of p is determined as p=0.001 which shows that there is a statistical significance between the groups. The novel Artificial Neural Network shows a better accuracy rate of 96.7100% using novel Artificial Neural Network than the K-means classifiers.