https://www.gov.uk/government/news/metropolitan-polices-use-of-facial-recognitiontechnology-at-the-notting-hill-carnival-2017 [Accessed 23rd. October 2017].
An investigation into biometric identifiers of the hand: A multimodal biometric for digital forensics Lilly Dan and Professor Sue Black
BIOMETRICS :
The statistical analysis of biological data ( OXFO R D E N G L I SH DI C T I O NA RY, 2018) .
In recent years, biometric technologies for personal identification have been applied in both commercial, security and civil contexts (Jain and Ross, 2015). Civil uses of biometric systems have now extended past the automation of DNA and fingerprint recognition, introducing image analysis biometrics for facial recognition into criminal justice process (Wiles, 2017).
MOTIVATION : A change in UK crime demographics has been attributed to the increased availability of technology, this has been expressed in the rise in the manufacture and distribution of indecent images of children (IIOC) (Home Office, 2016). The interpretation of evidence, for the verification of identity, is an element of digital forensics process which falls within the remit of a forensic anthropologist (Tully, 2016). The application of computer image comparison to this analysis would reduce observation error and provide a standardised process. It would also provide a platform to tackle issues regarding the use of biometric technologies in forensic application raised by government advisory bodies preceding the development of a national biometric strategy; due to be published June 2018 (Science and Technology Committee, 2018).
The hand is an appropriate anatomical interface as it encompasses a number of highly EVALUATION OF BIOMETRIC IDENTIFIERS differentiating anatomical features which have been frequently utilised by biometrics systems (Figure 1). In a forensically aware world, hands are more likely to be captured than Biometric Identifiers an individual’s face; this is especially prevalent in casework regarding IIOC carried out at 1. Epidermal Prints (Fingerprints) CAHId (Saini and Kapoor, 2016; Stratton, 2015). 2. Vein Pattern Analysis Nine anatomical features were identified and qualitatively evaluated as biometric identifiers using criteria developed for biometric systems (Jain et al., 2004) (Table 1). The addition of 3. Hand Geometry aetiology to the traditional criteria was included and justification for each decision was given. 4. Knuckle Creases Security, forensic and medically related biometric technologies were considered in the evaluation of general system performance. 5. Lunules 6. Scars 7. Tattoos 8. Skin Pigmentation Patternation
Figure 1. Multimodal biometric 9. Hair Patternation utilising the hand.
Table 1. Biometric suitability criteria with definitions. Adapted from Jain et al., (2004). The cause for the presence of the biometric identifier. An ideal biometric trait should be present in every one of the target population.
Collectability
Distinctiveness
A measure of how distinctive a biometric identifier is between individuals.
Circumvention
Permanence
Refers to the longevity and stability of the biometric trait.
Performance
Aetiology Universality
Acceptability
The ease in which a identifier can be captured and digitized. The public’s willingness to use the biometric. Whether or not the biometric trait can be easily spoofed enabling identity fraud or false identification An indication of the biometric system’s accuracy and speed.
ANALYSIS Observations regarding biometric identifiers • Identifier’s whose discriminative ability was more robust were generally of embryological aetiology. + • Pigmented marks and tattoos were highly distinctive. = • None of the identifier’s assessed satisfied the ideal biometric system requirements therefore a combination of several biometric identifier’s, a Figure 2. The evaluated outcome of a multimodal biometric system utilising epidermal ridges and multimodal biometric, would be suitable (Figure hand scars. 2). Issues identified • Biometrics and anatomical researchers have different motivations, many biometric Considerations in application to digital forensics systems have been created for security use, therefore the criteria used would be weighted Image analysis within digital forensics is challenging due to the differently with more emphasis on circumvention and performance. differing resolutions and positions in which a hand is captured. An • There is a fundamental misunderstanding amongst researchers regarding what appropriate biometric system would require a degree of flexibility constitutes a good anatomical identifier resulting in the misuse of the term ‘unique’ (Crisan, allowing the use of identifiers which are most appropriate to the 2017; Banerjee et al., 2017). context, a combination of identifiers for high resolution is • Commercial drive means that there is an oversaturation of systems and methods as well as presented in Figure 2. Advances within computer science image a lack of transparency regarding evaluation, testing and performance (Ngan et al., 2015). analysis such as segmentation and rotation can improve analysis • Population density of many anatomical identifiers is unknown. when hands are captured in an unstandardized position. • Forensic consideration of many identifiers have not been investigated and appropriate classification systems have not been verified.
CONCLUSION A multidisciplinary approach is needed to develop a multimodal biometric of the hand appropriate for application to digital forensics. A hierarchical multimodal structure with a base identifier of hand geometry would be appropriate for a bespoke approach towards different image contexts found within digital forensics (Kumar et al., 2014). Collaborative research and implementation of a multimodal biometric in a field of criminal justice will provide a natural framework for the improvement and standardisation of systems currently employed for forensic use in the UK (Crisan, 2017; Jain and Ross, 2015).
REFERENCES • •
• • • • • • •
• •
•
Banerjee, A., Basu, S., Basu, S. and Nasipuri, M. (2017). ARTeM: a new system for human authentication using finger vein images. Multimedia Tool Applications, pp 1-28. Crisan, S. (2017). A novel perspective on hand vein patterns for biometric recognition: problems, challenges and implementations. In Jiang R., Almaadeed S., Bouridane A., Crookes P., Beghdadi A. (eds) Biometric Security and Privacy. Signal Processing for Security Technologies. Springer, Cham, pp 2149. Home Office, (2016), Forensic science strategy: A national approach to forensic science delivery in the criminal justice system. Available at: https://www.gov.uk/government/publications/forensic-science-strategy [Accessed: 20/02/18]. Jain, A.K. and Ross, A. (2015). Bridging the gap: from biometrics to forensics. Philosophical Transactions of the Royal Society, B. 370: 20140254. Jain, A.K., Ross, A. and Prabhakar, S. (2004). An introduction to biometric recognition. IEEE Transactions on Circuit and Systems for Video Technology. 14 (1), 4-20. Ngan, M., Quinn, G. W. and Grother, P. (2015). Tattoo Recognition Technology – Challenge (Tatt-C) Outcomes and Recommendations. National Institute of Standards and Technology. Available at: http://nvlpubs.nist.gov/nistpubs/ir/2015/NIST.IR.8078.pdf [Accessed November 10th 2017]. Oxford English Dictionary. (2018). Biometric. Available at: https://en.oxforddictionaries.com/definition/biometric [Accessed 2nd January 2018]. Saini, M. and Kumar Kapoor, A. (2016). Biometrics in forensic identification: Applications and challenges. Journal of Forensic Medicine, 1(2), 1-6. Science and Technology Committee, (2018), Oral evidence: Biometrics strategy and forensic services, House of Commons. Available at: http://data.parliament.uk/writtenevidence/committeeevidence.svc/evidencedocument/science-and-technology-committee/biometrics-strategy-andforensic-services/oral/78113.pdf [Accessed 20/02/18]. Stratton, H. (2015). Dorsal hand feature analysis: an aid to forensic human identification. PhD dissertation, University of Dundee, Dundee. Tully, (2016), Method Validation in Digital Forensics Issue 1. Available at: https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/528123/FSR_Method_Validation_in_Digital_Forensics_FSR-G218_Issue_1.pdf [Accessed 20/02/18]. Wiles, P. (2017). Metropolitan Police’s use of Facial Recognition Technology at the Notting Hill Carnival, 2017. Available from: https://www.gov.uk/government/news/metropolitan-polices-use-of-facial-recognitiontechnology-at-the-notting-hill-carnival-2017 [Accessed 23rd October 2017].
