With the plethora of unmanned aerial vehicles (UAVs) deployed in Iraq, Afghanistan, and other hotbeds of radical militant fundamentalist activity worldwide, there's a wealth of opportunities to gather data. However, current techniques frequently come up short in imaging the low resolution fuzzy pictures taken by the drones and turning them into a usable piece of intelligence.
The Rochester Institute of Technology is stepping up to the plate to improve intelligence imaging techniques, and has won a $1 million Discovery Challenge Thrust grant from the Air Force Office of Scientific Research, which will finance the team's effort to design an entire package of improved imaging software and hardware to better track moving vehicles or individuals.
John Kerekes, associate professor in RIT's Chester F. Carlson Center for Imaging Science states, "The Air Force has clearly recognized the change in the threat that we have. I think we all understand that our military has a paradigm shift. We're no longer fighting tanks in the open desert; we're fighting terrorists in small groups, asymmetric threats."
The new imaging software will select the best imaging mode -- black and white imaging, hyperspectral or polarization -- depending on the conditions. It then begins collecting to threads of information -- one on the background, and the other on moving objects, particularly the one being tracked.
For simple images, black and white is used, which can track objects by shape. For objects of different colors, but harder to track by shape, hyperspectral imaging is used to take images of the objects in the color they appear at in multiple wavelengths, from the visible light to the near and short infrared parts of the spectrum beyond what the eye can see. The hyperspectral imaging distinguishes between virtually identically colored objects, indistinguishable to the human eye, thanks to slight variations in color.
However, should hyperspectral imaging come up short, the software has a third imaging technique up its sleeve -- polarization imaging. This imaging type measures surface roughness, helping even the best camouflaged targets to pop out from their surroundings.
Describes Professor Krekes, "These are all complementary pieces of information and the idea is that if the object you are tracking goes into an area where you lose one piece of information, the other information might help. The idea is to lead to more efficient sensing both from the point of view of collecting the data necessary and being able to adapt to these different modalities based on the conditions in the scene or the task at hand."
Zoran Ninkov, professor of imaging science at RIT, is designing hardware to feed information to Professor Krekes' software. The team is using tunable microelectronics devices to gather color information and a modified astronomical optical sensor for further imaging capabilities. Ohio-based Numerica Inc. is helping the software team design tracking algorithms. Together the team hopes to soon have a complete system, ready to help combat terrorist activity worldwide.
The software team is testing their initial models in a simulated environment similar to the popular online game Second Life. States Professor Krekes, "The idea behind that is you let the task at hand and the desire to optimize the performance drive what information is collected."