The Catholic University of America

Nov. 16, 2007

CUA Engineering Professors Prime Recipients of $2.7 Million Grant

Award Furthers Research on Computerized Camera System for Intelligence Community

Assistant Professor Scott Mathews and Associate Professor Mark Mirotznik, both from the department of electrical engineering and computer science, present their fly's eye imaging system.

Through its convex, compound eye, a fly creates an array of images, a panoramic view that enables the insect to detect motion 10 times faster than the human eye and so escape even the quickest flyswatter.

Now, thanks to a $2.7 million grant from the Office of the Director of National Intelligence, CUA Department of Electrical Engineering and Computer Science Associate Professor Mark Mirotznik and Assistant Professor Scott Mathews, along with a multidisciplinary, multi-university team of researchers, can continue developing an enhanced resolution digital camera that mimics a fly's eye view of the world.

The researchers have married the fly's eye imaging system to a computerized camera system, known as a Practical Enhanced Resolution Integrated Optical Digital Imaging Camera and dubbed PERIODIC. The PERIODIC system resembles a circuit-board sandwich, with the array of lenses on the front and green circuit boards in the middle carrying information to a computer.

Mirotznik explains, "The optical sensor and software components work in concert to solve underlying complex image registration and reconstruction problems in near real-time and produce high definition, multi-layer images."

In what the researchers call "computational imaging," 18 smaller camera lenses take sub images and then a computer combines their information to produce a single, enhanced image with an increased resolution.

Mirotznik offers the example of a taking a picture of a person in front of a brightly lit window. A conventional camera wouldn't know how to adapt to the changes in light, and therefore would not be able to properly capture most of the details. The fly's eye camera, however, can capture even the smallest details in that scene.

"It's not meant to make a picture more pleasing to the eye, but to enable a computer to put together more information than the eye (or the typical camera) can see," says Mirotznik.

The prototype, originally funded by a U.S. Disruptive Technology Office challenge grant, greatly improves the resolution and dynamic range of imagery, removes glare, and performs spectral filtering. This new grant will help fund an additional 24 months of development, paying for more equipment and graduate student assistants.

"We're very pleased that this is such a huge research effort and now we've got a lot of work ahead," Mathews says.

The camera is being developed for purposes including iris recognition. Existing technology can scan an iris, which the researchers liken to a fingerprint, to identify a person. But it is not advanced enough to scan noncompliant subjects who may not be at close range and do not maintain eye contact with the camera for a length of time. This enhanced camera technology aims to solve that problem, a huge boon for the intelligence community.

Miroztnik and Mathews are particularly gratified by application of the technology to burn injury assessment. "The most critical factors determining whether or not burn patients recover are rapid assessment of the degree of burns and quick, appropriate treatment," says Miroztnik.

"Using multi-spectral imaging, we can accurately measure blood oxygen levels in a patient's skin and give doctors the rapid evaluation of the extent of tissue damage they need to initiate appropriate treatment."