John Daugman correctly summarized the state of forensic iris recognition circa 2006 for the book Forensic Human Identification: an Introduction by Thompson (CRC Press, 2006). Iris recognition has limited forensic value, because (unlike fingerprints or DNA, for example) (1) iris patterns are not left behind at crime scenes; (2) and in death the pupil usually dilates significantly, the cornea clouds, and the iris tissue degrades relatively rapidly. (3) Moreover, currently available iris databases are quite small (only a few million digitized samples of iris patterns exist today); and because of the novelty of this biometric, (4) such data currently has no legal or established forensic status as admissible evidence. [Numbers () added.] In the intervening ?15 years, all of Daugman s observations, save one, have been overtaken by events: (1) The advent of ubiquitous high resolution video/photography has led to widespread collection/retention/dissemination of imagery of sufficient resolution for iris recognition. (2) Demonstrations of post-mortem iris recognition have been made. (3) Large iris databases have been constructed. The last issue (4) regarding admissible evidence remains to be resolved. Forensic iris was a topic at the June 2018 Iris Experts Group Meeting1. Key issues discussed there were: measurements and analysis that need to be done to provide the underpinnings for a resolution of the forensic status of iris recognition and the development of documentation for such measurements and analysis that will enable explanation of iris collection and recognition to lay audiences, including those in a courtroom. An important point was that the perceptions of the public and the popular media with respect to biometrics and to iris recognition in particular are frequently inaccurate and must be considered in any development of materials designed to explain iris recognition to the lay public. To help resolve the questions discussed at that meeting, this paper reviews the current state of the art in iris recognition, the perceptions of the public regarding iris recognition, and makes suggestions regarding measurements and analysis that will help enable use of forensic iris in appropriate settings going forward. We welcome comments for the next revision of this document. Please send comments to james.matey@nist.gov.
原文链接
分享
引用
