Research gets hands-on with fingerprint evaluation

Three inexpensive programs could help determine the quality of fingerprints far better than what humans are capable of, researchers at Penn State say.

Elizabeth Armstrong Moore
Elizabeth Armstrong Moore is based in Portland, Oregon, and has written for Wired, The Christian Science Monitor, and public radio. Her semi-obscure hobbies include climbing, billiards, board games that take up a lot of space, and piano.
Elizabeth Armstrong Moore
2 min read
The fingerprint on the left was prepared using an older development technique; the image on the right, revealing more ridge detail, is the right half of the same fingerprint and was prepared using the new technique. Akhlesh Lakhtakia/Penn State

When it comes to evaluating fingerprints, not everyone has the keen eye of Sherlock Holmes. In fact, the human eye has proven to be subjective -- and thus not fully reliable -- when it comes to determining the mere quality of the print.

So researchers at Penn State are hoping to remove that subjectivity with the merging of three inexpensive computer programs that grade that quality -- which can be affected by all sorts of environmental weathering and smudging -- on a scale of 0 to 100.

"Humans can't grade finer than the zero to three scale, but computers can," Akhlesh Lakhtakia, a professor of engineering science and mechanics at Penn State, said in a school news release. "2.3 percent is worse than 15 percent, but both could be graded as a zero by the naked eye."

Reporting in the current issue of the journal Forensic Science International, Lakhtakia and his team describe their novel approach that uses the FBI's Universal Latent Workstation, the open-source image editor GIMP, and a custom pixel-counting program written in Mathematica.

First, a photo of the print is run through the Workstation, which creates a map of the fingerprint that shows the background in black, areas with definite ridges in white, and less obvious regions in yellow and blue.

That map is then run through the GIMP editor, which converts it into an image file with red-green-blue values that are stored as number clusters to be translated into binary sequences.

Finally, the pixel-counting algorithm calculates the percentage of white pixels, which is then converted into a 0 to 100 grading scale -- 100 being a high-quality fingerprint with the most number of white pixels.

"The next step of this kind of research is, is there false detail created by development techniques?" Lakhtakia asked. "That can happen. Looking at the thin-film technique that my group has developed, I don't imagine so, but we would obviously have to prove it."

A subjective grading system on a scale of 0 to 100 could conceivably come into play in court if, say, a defendant is being tried using as evidence a fingerprint that has a confidence grade of 2 out of 100. That would ideally hold more weight than a forensic expert judging the quality as iffy.

The U.S. Department of Justice funded the research.