Iris patterns as a Biometric tool for Forensic Identifications: a Review

: Human biometric features are considered unique, permanent, measureable and objective traits used as authentic tools for identification of an individual. Biometric authentications are increasingly utilized for identifying a genuine user or an imposter. The automated iris pattern recognition is based on the unique, non-intrusive and un-imitable micro-features of human iris. Several mathematical and statistical algorithms have been suggested for extracting and encoding iris minutiae used for pattern recognitions in automated systems. Iris scanning is an easy and non-invasive technique, having comparatively lower false acceptance and rejection rates and negligible chances of manipulations and spoofing. Despite of some limitations, iris recognition is considered a promising advancement in forensic sciences. Present review has presented the iris recognition knowledge in nutshell to the readers.

important field of biometrics which can authorize or authenticate a person from an imposter effectively. Biometrics can provide the most accurate and highly secured identification and verification systems with multiple applications 2 . The authentication via biometric verification is becoming increasingly common in security applications such as homeland security, banking, border control, access control, web-based services, welfare distribution schemes, forensics etc. Several biometric modalities have been suggested for the secured forensic applications, though each one has its own advantages and limitations. No single biometric feature can meet all the performance requirements of a system 3; so use of multimodal biometric systems has been advocated to overcome the limitations of unimodal means of identification. The emergence of biometry and the computer technology is taken to be contemporary during second half of twentieth century. Rapid developments in computer technologies, computational capabilities and computing approaches have fantastically improved upon the image capturing process, feature extraction, feature robustness, and feature comparisons 2 .
Biometric features of a person are very difficult to be forged, imitated, lost, transferred or stolen and the security systems based on such human traits require the presence of a genuine user for getting access to a particular resource.
Biometric identity signatures include some unique physical, physiological and behavioral attributes, also known by the names like biometric traits, indicators, identifiers or modalities. Physical attributes of biometry include some prints (fingerprints, footprints, lip prints, ear print), scans (facial scan, retinal/iris scan, MRI/CT scan), body odors, vein-networking etc., whereas the behavioral elements may include the voices, handwritings, signatures, typing patterns, gait patterns etc. Biometric authentications have been widely utilized by a number of government (border crossing, airport security, passport control, welfare distribution schemes, driving licenses), commercial (access control, ID cards, e-commerce, cellular phones, credit cards) and investigative/forensic Orlando shootings and Brussels airport explosions), have necessitated the introduction of some more advanced security and surveillance systems to replace the traditional and obsolete methods 4 . The uniqueness, permanence, measurability, performance, acceptability, universality, and low circumvention are the important characteristics of a biometric trait to make it suitable for forensic identifications 5 , and Iris pattern recognition is one of forensic modality having all these characteristics which can work even in nonideal situations 3 .
Present review article has been written with an intention to provide an in-depth and nutshell overview of iris recognition as a biometric tool of recognition to the readers.
The conclusive information presented in some recently published studies was integrated to present the current status of iris pattern recognition as a forensic tool. The related articles published in some journals and books were searched through online search engines like Pubmed, Scopus, Embase, Web of Science etc., and, the information regarding uses, advantages, limitations, algorithms developed, and future probabilities of iris pattern recognitions and biometry as a whole were extracted out to present this review.

Iris recognition as a forensic tool
The ocular region of human eye possesses the most accurate, highly reliable, well protected, fairly stable and almost unforgable biometric signatures like iris, retina and sclera vein patterns; the iris being the most significant among them [6][7] . Iris recognition is based upon its unique and stable textured features like crypts, furrows, nevi, corona and freckles [7][8] . It analysis the random patterns of the iris by mathematical pattern recognition algorithms or techniques. The automated iris pattern recognition is among the most recent biometric methods having tremendous potential as an infallible mode of personal identification, especially at high security areas like airports, nuclear reactors, embassies, parliaments, immigration controls etc. Iris patterns can also be used in

Iris recognition process and its advantages
The basic iris recognition process involves four steps i.e., image acquisition, iris feature extraction, iris pattern matching and digital representation [11][12]15 . Acquisition involves capturing the full digitalized image of the eye, preferably from a distance ranging from 10 cm to 1 m. The visual and infrared lights used in camera help in capturing unique minute details of iris and in isolating the iris from pupil, respectively; though the combination of these two types of lights are suggested to provide much better results 18,24 . Iris patterns as a forensic tool have an edge over the other means of biometric identifications due to certain reasons. Iris scanning is an easy and non-invasive technique (as iris remains well protected within layers), have lesser chances of its manipulations or spoofing 9,11,15 . Iris patterns have small intra-class variability and high degree of randomness. Unlike retina, iris does not get decomposed immediately after death and hence can be utilized for corpse identification up to certain hours after death.
Trokielewicz et al 25 reported that more than 90% iris images captured few hours after death (under mortuary conditions) were still correctly recognizable and the serious iris deterioration begun approximately 22-hours post-mortem. Thus, human iris can be successfully used for biometric authentication even after death, though within certain limited period. The richness of texture details in iris images enhances its suitability for forensic identifications 15 . Iris patterns don't get altered even after surgical operation of eye; and also the use of eyeglasses or contact lenses do not interfere in iris pattern recognitions 11,14,18 . Iris matching is possible even for blind people having their iris intact 18 . Researchers have shown that the iris scans have comparatively low false acceptance and rejection rates 14,22 , so iris methods are preferred for security management of highly secured areas.

Limitations of iris patterning as a biometric modality
Iris, itself, is a very small organ to be scanned from a distance. Iris recognition is a challenging task to be performed with desired accuracy from a distance larger than a few meters. Some factors which may discourage scanning process for iris pattern recognition include the illumination and contrast differences, inadequate image quality, occlusion of eyelids and eyelashes, non-linear deformations, rotation differences, defocus, reflective surfaces, stop and stare interface of cameras, advancing age changes like drooping eyes etc. 8-9, 11, 121-13, 23, 26 . Intensive exposure to environmental contaminants like pollutants, metal vapors, smoke etc., may affect iris pigmentations and its discrimination capabilities 11 . Diseases like cataract, iritis, iridocyclitis, diabetes etc., may also affect iris scanning 13 . Some medical and surgical procedures often affect the overall shape and colour of an iris; however, the fine structure details of iris remain intact for many decades. Iris recognition requires extremely dedicated arrangements for imaging and sufficient cooperation from the user/subject. Venugopalan and Savvides 27 reported that development of spoofed iris, accurately identical to actual iris, is also possible. To make iris recognition a human-friendly technique (especially for forensic purposes), the similarities between irises should be made visualizable, interpretable and explainable 28 . Benaliouche and Touahria 22 found that iris and fingerprints have the same matching speed and accuracy, however, multimodal approach may give better results for recognition process with lower errors. Inspite of various limitations and problems discussed above, iris scan may be taken as a promising advancement in the forensic contexts as well in industry for commercialization purposes.

Conclusions
The iris pattern recognition is relatively a very recent biometric modality that can contribute significantly in forensic identifications of an individual. Human iris patterns are unique, permanent, un-imitable, measureable and universal features that ensure their use as an objective biometric modality. The unique features of iris make it one of the most secured methods of authentication. There are some advantages and inherent difficulties using iris patterns as infallible mean of identification, however, the use of iris patterns for authentication and identification purposes has a tremendous scope in forensic sciences. The scientific community is continuously involved in research to develop a universal iris-based algorithm that may help its unanimous acceptance as a biometric tool for authentication and identification. Iris patterns remain stable for few hours after death and hence can be valuable adjunct for identification of a recently dead person, if other features don't allow the same. Being a comparatively newer technology, extensive testing and research is required before confirming use of iris patterns at