3D-Finger knuckle Recognition using Convolutional Neural Network
Dua’a Hamed Al-Janabi, Ali Mohsin Al-Juboori · 2022
In the previous few years, a very massive development has happened in biometrics methods by artificial intelligence tools, especially in the recognition field. Unique of the essential things is a contactless three dimension (3D) finger knuckle pattern. 3D images of finger knuckles encompass more information that can accommodate the development of biometric recognition. Performance attacks can also be easily discovered by checking the three dimension (3D) images. So, this paper motivates the progress of this biometric identifier which can suggest an accurate, effective, and convenient other for biometric recognition. In this paper, the three_dimention image has been constructed from a series of two-dimensional (2D) images, and also a new model has been examined based on the convolutional neural network (CNN). The proposed model has used The Hong Kong Polytechnic University (PolyU) three dimension (3D) Finger Knuckle Images datasets. The attained classification accuracy was 98%. The offered classification model offers an active and active system for evaluating finger knuckle classification with high dependability and accuracy.