A CNN and VGG- 16 Approach for Identification System Based on Ear Biometrics

Abstract

cAbstract: With the increase in digitization over the entire world, there is a growing concern for security and privacy. Alphonse Bertillon, who created a technique to identify a person by precise body measures, is credited with developing biometrics for the first time in Paris during the 1800s. Since then, biometrics have greatly aided in preserving human security and privacy. The proposed study offers a through investigation of the ear as a biometric system utilising a novel taxonomy. It also provides a detailed analysis of databases, performance evaluation criteria, and cutting-edge methodologies. This research work acquired 99.8% accuracy and 0.2% loss while training and testing the model. Convolution Neural Network (CNN) is used due to its proven accuracy in object detection. To achieve better results in large scale image recognition. The research has been supported with the datasets from Kaggle. As a benchmark for comparison, the database from IIT Delhi. The results show an enhancement over the existing techniques and act as a springboard for new ear biometrics researchers.