Infrared technology could help the military to remotely and safely identify sites where nuclear weapons are being made.

Researchers at Brigham Young University have developed a model that distinguishes the material in every pixel of an image taken with a long-wave IR camera. Using data from Lawrence Livermore National Laboratory, they used infrared images to analyze more basic materials.

This new technique also can group together related pixels that to map out the various materials in an image.

"What we wanted to know is if you didn't know anything about the material in an image, and we had a number of pictures over time, could we let the algorithms figure out what the different materials are and separate them out," said Gustavious Williams, professor of engineering at BYU.

In the study, incoming signals were directly separated so the material’s unique signature could be identified. Achieving this has proved problematic in the past, as numerous other incoming signals, such as an object’s temperature and surrounding weather conditions, add noise and cloud the material’s light signature when hyperspectral IR cameras are used.

What’s unique about the BYU study is that it determines the material’s unique signature by separating the incoming signals, whereas other methods handle noise by matching the combined signals in a database. The resulting information is similar to measuring the material with a spectrometer in a lab.

"If we apply this model, we can get distributions on the physical characteristics of each of these pixels and, using those different characterizations, also cluster the pixels with like materials," said Candace Berrett, a statistics professor at BYU.

The researchers are continuing to develop the IR imaging and analysis technique, which they expect will enable remote detection of exact materials, chemicals and gases being received and emitted from sites of suspected illegal nuclear production.

Besides nuclear weapons detection, the method has potential for applications such as spotting chemical and gas leaks in areas that have been struck by an earthquake, tornado or other natural disaster.

"There are already instruments that can do this,” Williams said. “Our algorithms take a different approach but are still at a basic research stage. There are lots of places this research could go."

The work was funded by a grant from the US National Nuclear Security Administration. The research was published in Technometrics (doi: 10.1080/00401706.2013.869262).

For more information, visit: www.byu.edu