An Effective Iris Identification Software System for Hospitals’ Emergency Unit
Yoan Nikaros Suwardi, Mir Shahriar Emami · 2024
Iris patterns are unique and immutable, ensuring reliability in terms of individuals identifications. Previous methods are time-consuming, potentially compromising patient care. This work addresses the imperative for efficiency in healthcare, particularly in patient registration during emergencies. This paper has proposed an effective leveraging iris recognition software system for swift and accurate identification of patients in emergencies. In this paper, we used Hough transform for segmentation in localizing iris and pupil regions. The segmented iris region was then normalized to a rectangular block with fixed polar dimensions using Daugman’s rubber-sheet model. To extract features from the normalized iris image, which was at a constant size, we performed PCA combined with DWT for analyzing the data matching rate among iris templates. The experiments conducted through CASIA-IrisV4 dataset and iris images were captured by personal handphone camera, then stored, and mapped for seamless matching and verification. The highly accurate results (97.96%) obtained through experiments showed that the software system proposed in this research can be successful to assist emergency units in hospitals to identify patients in emergency and record their identification information as soon as they arrive at the hospital.