Palm Print Recognition using Deep Learning

In recent decades, numerous studies have focused extensively on biometric palmprint recognition. Palm print recognition has gained significant popularity and importance across various domains owing to its exceptional efficiency and accuracy in personal identification. The biometric characterization of a person's palm print is unique. However, a way to enhance the image is needed in order to produce a better and clearer image. Recently, palm print recognition methods based on features acquired using a series of convolutional neural networks have been introduced, among which DenseNet-121 has a densely connected structure, unlike other structures. This paper presents a scheme for palm print recognition by image enhancement. Contrast-limited adaptive histogram equation (CLAHE) is one of the image enhancement methods that can provide bounded segment and region size and is based on deep learning using DenseNet-121. To measure performance, the CASIA dataset was used. Experimental results on the DS show that the palm print features of Denes 21 achieve a recognition accuracy of 99 %, demonstrating the effectiveness and reliability of the proposed palm print.