Deep Learning-Based Biometric System Analysis of Palmprint Images
Mohammed Jaafar Rashid Al-Majmaie, Mesüt Çevik · 2022
The implications, both favorable and negative, can result from the speed with which technology is progressing. An increasing number of serious crimes are being committed in cyberspace, and as a result, there has been a need for critical research into the topic of information security all over the globe. Therefore, biometric recognition systems are being used by so many businesses and government agencies today. This study applied a novel strategy for human identification based on a convolutional neural network pre-trained model called AlexNet and the wavelet transform. The suggested method took photographs of people’s hands and extracted novel and effective characteristics from them, then fed those features into an ensemble learning classifier, which used those features to divide the images into many classes, each of which represented one of the 72 people. The proposed method combined Alexnet pre-trained model combined with wavelet transform and ensemble learning. When compared to recent studies at the cutting edge, the 99.14 success rate obtained here is outstanding