Prayer Sitting .. Is it a Biometric?
Bassam M Al-Mahadeen, Rufaydah Alhuweimel, Khalid Mohammad Jaber · 2023
sometimes the available data and its nature determine us in choosing what biometric to rely on for identifying people. From here, we thought outside the box in a mechanism for building the biometric that we can rely on in order to overcome the challenges that arise in human identification. So, in this research paper we find a way to link between foot geometry with the way of sitting in prayer and adopting it to distinguish between persons. We used a Convolutional Neural Network (CNN) algorithm on 900 images of different persons while sitting to perform the prayer; the collected images were preprocessed using different sharpening filters then making segmentation to highlight fine details and enhance edges, then divided into two groups: training set and testing set to be inserted into an algorithm to extract unique characteristics from them and train the CNN on them, store as a template to be compared in testing cases. We obtained a result that the prayer sitting is biometric with an accuracy of 91.2% based on the data set that was used.