Authorities say no two fingerprints have ever been found identical among the many billions of human and automated computer comparisons done to date. They're the most common tool for the identification of people with a criminal history in nearly every police agency around the world, and they're the single most commonly gathered forensic evidence. Fingerprint evidence alone easily outnumbers all other forensic examination data combined. Tens of thousands of people have their fingerprint impressions added to forensic databases each day.
It's also acknowledged that automated systems to identify fingerprints can prove erroneous from 1 to 4 percent of the time.
It's thought that a Persian book from the 14th century, Jaamehol-Tawarikh, included the first known mention of the practice of identifying individuals from their fingerprints. But it wasn't until 1880 when Dr. Henry Faulds used printer's ink to make the first fingerprint identification of a greasy impression on a bottle of alcohol that the technique took hold.
But it was the work of Argentine policeman Juan Vucetich which made the first criminal fingerprint identification possible in 1892. Vucetich identified one Francis Rojas, a murderous woman who dispatched her two sons and then cut her own throat to avoid suspicion, as the culprit. She was undone by the discovery of her bloody fingerprint on a door post, and that evidence was used to mark her as the murderer.
Now Anil Jain, a former winner of the Guggenheim Fellowship from the Department of Computer Science and Engineering at Michigan State University, and his colleagues have developed a method of taking a two-dimensional image of a fingerprint and mapping it directly to a 3D printed surface. Jain's team includes Sunpreet Arora, a computer science doctoral candidate, Kai Cao, a research associate in computer science and engineering and researcher Nick Paulter from the National Institute of Standards and Technology.
The output is replete with all the whorls and ridges of an actual human fingerprint, and Jain and his team call them 'fingerprint phantoms.'
In 1924, the US Congress established the Identification Division of the FBI. The IACP's National Bureau of Criminal Identification and the US Justice Department's Bureau of Criminal Identification were consolidated to form a common repository of FBI fingerprint files, and by 1946, the FBI had processed 100 million fingerprint cards and kept them on file. By 1971, that number rose to 200 million cards.
Then came the automated fingerprint identification system, or AFIS, a technology to computerize criminal files and manually maintained fingerprint files. Now the single largest AFIS repository in America is operated by the Department of Homeland Security's US Visit Program, and it contains the fingerprints of more than 120 million people.
"A 3D heart or kidney can be created," Jain says. "Because the dimensions are known, they can be put into a scanner and the imaging system can be calibrated. When I have this 3D fingerprint phantom, I know its precise measurements, and because I know the true dimensions of the fingerprint features on this phantom, I can better evaluate fingerprint readers."
While Jain says he plans to use the technology to improve the accuracy of fingerprint-matching and biometric systems, it could have a variety of other uses – and misuses – as well.
"Government agencies worldwide encounter individuals who have gone to extreme measures to alter their fingerprints to avoid being identified by Automated Fingerprint Identification Systems (AFIS)," Jain says. "We have developed an algorithm capable of automatically looking at the fingerprint image and determining whether the pattern is 'unusual' or 'normal.' If it is 'unusual,' then it's flagged as a potential altered fingerprint and the suspect's fingers will be manually examined for signs of alteration."
But according to Jain, that still leaves some room for variation and uncertainty.
"Of course, it is not possible for us to determine whether the observed unusual fingerprint pattern was due to an injury to the finger or due to intentional mutilation of the fingerprint."
Jain says that new fingerprint scanners are tested by running millions of fingerprint images through systems software, but as the 2D images aren't as accurate as 3D versions of fingers, it's possible that discrepancies go unnoticed during testing.