Forensic Iris Recognition: A Survey.

Iris recognition is a biometric technology that utilizes the feature sets in an individual's iris for identification purposes. Iris recognition is a non-invasive technique that does not require physical contact with the identified individual. Post-mortem iris recognition refers to using iris images from a deceased person to identify or verify their identity. It has several potential applications, such as forensic investigations and disaster victim identification. However, biological changes after death can cause significant differences between the post-mortem and ante-mortem iris data, which presents challenges for iris-capturing sensors, feature extractors, and iris matchers. This paper surveys existing research on using iris images for post-mortem identification, including a comprehensive review of the state-of-theart and a summary of the latest results and observations. This survey has several unique elements, which provide a valuable resource for researchers and practitioners seeking to understand the capabilities and limitations of post-mortem iris recognition technology. Firstly, we discuss the post-mortem iris recognition steps and biological changes in the iris texture after a person's death from a medical standpoint. We then present the approaches to address the post-mortem iris recognition problem, including traditional iris recognition techniques, deep learning-based strategies, and interpretable methods. Furthermore, we provide the potential confounding factors that might impact the recognition performance. We also offer a comprehensive review of the publicly available post-mortem iris databases and the evaluation metrics used to assess the performance of post-mortem iris recognition systems. Finally, we conclude the paper by providing a constructive discussion and emerging future research directions. [ABSTRACT FROM AUTHOR]