Abstract
The fingerprint identification is the most widely used authentication system. The fingerprint uniqueness for each human being provides error-free identification. However, during the scanning process of the fingerprint, the generated image using the fingerprint scanner may differ slightly during each scan. This paper proposes an efficient matching model for fingerprint authentication using deep learning based deep convolutional neural network (CNN or ConvNet). The proposed deep CNN consists of fifteen layers and is classified into two stages. The first stage is preparation stage which includes the fingerprint images collection, augmentation and pre-processing steps, while the second stage is the features extraction and matching stage. Regarding the implantation results, the proposed system provided the best matching for the given fingerprint features. The obtained training accuracy of the proposed model is 100% for training dataset and 100% for validating dataset.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have