Engineers American Universities of Columbia and Buffalo created a new analysis artificial intelligence fingerprint (AI) What reset a long-standing belief in forensic science that no two fingerprints are alike, not even on different fingers of the same person.
The discovery, which was published this Wednesday in the journal Science Advance magazine
demonstrated with There is a 99.99% chance that the fingerprints of any two fingers from the same person are much more similar than previously thought..
Fingerprints are required in crime labs to solve cases, and on billions of mobile phones around the world for digital authentication..
Now the research is led Gabe Goengineering student at Columbia University, along with other researchers from the same university and from Buffalo, showed that this limitation can be overcome by analyzing different characteristics traces that have not yet been taken into account.
Students found A public US government database containing approximately 60,000 fingerprints.some from the same person and others from different people.
Engineers, They extracted representation vectors from traces. and they watched some amazing results: fingerprints from different fingers of the same person are extremely similar.
They discovered that the key was in ridges (the most visible area of the trace), since it orientation near the center of the tracks largely explained this similarity and the fact that this pattern was repeated in all pairs of fingers of the same person
.