New+technology+detects+altered+fingerprints

Fingerprint Recognition

INTRODUCTION Michigan State University has licensed innovative software that detects altered fingerprints to Morpho, part of the Safran group, one of the world’s major suppliers of identification and detection solutions. Fingerprint recognition has been successfully used by law enforcement agencies to identify suspects and victims for almost 100 years. Recent advances in automated fingerprint identification technology, coupled with the growing need for reliable person identification, have resulted in an increased use of fingerprints in both government and civilian applications such as border control, employment background checks, and secure facility access. Examples of large-scale fingerprint systems in the US government arena include the US-VISIT’s IDENT program and the FBI’s IAFIS service. The success of fingerprint recognition systems in accurately identifying individuals has prompted some individuals to engage in extreme measures for the purpose of circumventing these systems. The primary purpose of fingerprint alteration is to evade identification using techniques varying from abrading, cutting, and burning fingers to performing plastic surgery. The use of altered fingerprints to mask one’s identity constitutes a serious “attack” against a border control biometric system since it defeats the very purpose for which the system was deployed in the first place, i.e., to identify individuals in a watch list. It should be noted that altered fingerprints are different from fake fingerprints. The use of fake fingers—made of glue, latex, or silicone—is a well-publicized method to circumvent fingerprint systems. Altered fingerprints, however, are real fingers that are used to conceal one’s identity in order to evade identification by a biometric system. While fake fingers are typically used by individuals to adopt another person’s identity, altered fingers are used to mask one’s own identity.

Photographs of altered fingerprints. (a) Transplanted friction ridge skin from sole. (b) Fingers that have been bitten. (c) Fingers burnt by acid. (d) Stitched fingers. 2 BACKGROUND Fingerprint alteration has a long history. As early as 1933, Gus Winkler, a murderer and bank robber, was found to have altered the fingerprints of his left hand except for the thumb by slashing and tearing the flesh of the fingers. Further, the pattern type of one finger was altered from double loop to left loop. In more recent cases, a man using the name Alexander Guzman, arrested by Florida officials in 1995 for possessing a false passport, was found to have obfuscated fingerprints. After a two-week search based on manually reconstructing the damaged fingerprints and searching the FBI database containing 71 million records, the reconstructed fingerprints of Alexander Guzman were linked to the fingerprints of Jose Izquierdo who was an absconding drug criminal. His fingerprint mutilation process consisted of three steps: Making a “Z” shaped cut on the fingertip, lifting and switching two triangular skin patches, and stitching them back. In September 2005, a drug dealer named Marc George was apprehended because his limping gait as a result of surgery caught the attention of border officials. It is not just the criminals who have been found to alter their fingerprints. In December 2009, a woman successfully evaded the Japanese immigration AFIS by surgically swapping fingerprints of her left and right hands. Although she was originally arrested for faking a marriage license, scars on her hands made the police suspicious. Fingerprint alteration has even been performed at a much larger scale involving a group of individuals. It has been reported that hundreds of asylum seekers had cut, abraded, and burned their fingertips to prevent identification, by EURODAC, a European Union-wide fingerprint system for identifying asylum seekers. Although the number of publicly disclosed cases of altered fingerprints is not very large, it is extremely difficult to estimate the actual number of individuals who have successfully evaded identification by fingerprint systems as a result of fingerprint alteration. Almost all the people identified as having altered their fingerprints were not detected by AFIS, but by some other means. Vulnerability of Fingerprint Identification Systems Fingerprint alteration is a serious threat to AFIS since it revokes one of the fundamental premise that fingerprint is persistent during one’s lifetime. To understand the vulnerability of AFIS to fingerprint alteration, there have been many test that used a commercial matcher, VeriFinger SDK 4.2, to match 1,335 altered fingerprints to their mated pre-altered fingerprints. To establish a baseline, NIST SD4 database, which consists of 2,000 fingers with two impressions per finger, was used to obtain genuine and impostor match score distributions using VeriFinger SDK. 3 The following images shows the score distributions for pre/post-altered fingerprint pair matches according to type and genuine and impostor matches in NIST SD4. The key observations here are: 1. The match score distributions of pre/post-altered fingerprint pairs for all alteration types follow the impostor score distribution. 2. Heavy tails in pre/post-altered match score distributions indicate that fingerprint alteration, as observed in our database, is not always successful in evading AFIS. 3. At a threshold of 41, which corresponds to 0 percent False Acceptance Rate (FAR) on NIST SD4, 83 percent of the pre/post-altered fingerprint pairs have genuine match scores below the threshold. This means that AFIS is unable to link most of the altered fingerprints to their true mates. The image below show’s examples where altering a fingerprint leads to failure in matching to its true mate. The process of fingerprint mutilation destroys the ridge structure itself so that minutiae extraction is not possible in this area. Also, severe ridge distortion, such as ridge structure transformation or ridge deformation due to scars, alters the spatial distribution of the minutiae. There is no guarantee that fingerprint alteration will always be successful in evading AFIS. As long as there are a sufficient number of minutiae that can be extracted in the unaltered area, pre/post-altered fingerprint mates can be successfully matched.

Examples where fingerprint alteration severely degrades the matching score with the pre-altered mates. (a) Mutilation over a large area. (b) Ridge transformation. These altered fingerprints have a match score of 0 with their true mates. All squares indicate minutiae extracted from the image and squares filled with red color represent matched minutiae between the pre/post-altered fingerprints. Examples where the pre/post-altered fingerprint mates are correctly matched despite fingerprint alteration. (a) Alteration with a small damaged area and no ridge distortion. (b) Sufficient number of minutiae in the unaltered area even with severe fingerprint alteration. Only a few corresponding minutiae are connected with dotted lines.

CONCLUSIONS AND FUTURE WORK The success of AFIS and their extensive deployment all over the world have prompted some individuals to take extreme measures to evade identification by altering their fingerprints. The problem of fingerprint alteration or obfuscation is very different from that of fingerprint spoofing, where an individual uses a fake fingerprint in order to adopt the identity of another individual. While the problem of spoofing has received substantial attention in the literature, the problem of obfuscation has not been addressed in the biometric literature, in spite of numerous documented cases of fingerprint alteration for the purpose of evading identification. While obfuscation may be encountered with there biometric modalities (such as face and iris), this problem is especially significant in the case of fingerprints due to the widespread deployment of AFIS in both government and civilian applications and the ease with which fingerprints can be obfuscated. Since the NFIQ has limited ability in distinguishing altered fingerprints from natural fingerprints, we developed an algorithm to automatically detect altered fingerprints based on the characteristics of the fingerprint orientation field and minutiae distribution. The proposed algorithm based on the features extracted from the orientation field and minutiae satisfies the three essential requirements for alteration detection algorithm: 1) fast operational time, 2) high true positive rate at low false positive rate, and 3) ease of integration into AFIS. The proposed algorithm and the NFIQ criterion were tested on a large public domain fingerprint database (NIST SD14) asnatural fingerprints and an altered fingerprint database provided by a law enforcement agency. At a false positive rate of 0.3 percent, the proposed algorithm can correctly 66.4 percent of the subjects with altered fingerprints, while 26.5 percent of such subjects are detected by the NFIQ algorithm. This study can be further extended along the following directions: 1. Determine the alteration type automatically so that appropriate countermeasures can be taken. 2. Reconstruct altered fingerprints. For some types of altered fingerprints where the ridge patterns are damaged locally or the ridge structure is still present on the finger but possibly at a different location, reconstruction is indeed possible. 3. Match altered fingerprints to their unaltered mates. A matcher specialized for altered fingerprints can be developed to link them to unaltered mates in the database utilizing whatever information is available in the altered fingerprints. 4. Use multibiometrics to combat the growing threat of individuals evading AFIS. Federal agencies in the United States have adopted or are planning to adopt multibiometrics in their identity management systems (the FBI’s NGI and DoD’s ABIS). However, other biometric traits can also be altered successfully. It has been reported that plastic surgery can significantly degrade the performance of face recognition systems and that cataract surgery can reduce the accuracy of iris recognition systems. To effectively deal with the problem of evading identification by altering biometric traits, a systematic study of possible alteration approaches for each major biometric trait is needed.

http://www.youtube.com/watch?v=tnueMG7cDJc&feature=youtu.be