RIT hosts image processing workshop Friday

‌

Nathan Cahill

Data, data, everywhere…

Digital image processing is essential in an era of abundant data. Computer algorithms search multiple images for specific information that is layered and compressed to create information rich “maps” of tumors, disaster zones, people and blue cars (or other such objects).

Scientists from around the region will share their image-processing research at the Western New York Image Processing Workshop in Louise M. Slaughter Hall on the Rochester Institute of Technology campus on Nov. 22. The event is sponsored by the Rochester chapter of the IEEE Signal Processing Society with cooperation from the local chapter of the Society for Imaging Science and Technology.

Highlights will include a keynote lecture on forensic bitemark comparisons by H. David Sheets of Canisius College and a lesson in searching for people in images by alumnus Andrew Gallagher ’00 (MS, electrical engineering) of Cornell University. Kyros Kutulakos, from the University of Toronto, will also present a talk on “Imaging Less than Meets the Eye.”

“Hosting this workshop builds on RIT’s reputation of being a premier location for imaging research,” says Nathan Cahill, associate professor in RIT’s School of Mathematical Sciences and conference chair.

The RIT organizing committee includes Behnaz Ghoraani, assistant professor of biomedical engineering; Raymond Ptucha, assistant professor of computer engineering; and Survi Kyal ’13 (MS, electrical engineering), research engineer from Xerox Corp.

The daylong event will include paper and poster presentations by RIT students and scientists as well as researchers from the George Eastman House International Museum of Photography and Film, State University of New York at Fredonia, University of Rochester Medical Center and Xerox, among others.

“The event provides a great opportunity for academic and industrial researchers in the region to come together to share their research,” Cahill says. “For students—both those presenting papers and those who just attend—the workshop provides a snapshot of research in the region and also allows them to network with people from local companies and other universities who work in areas related to image processing.”

Tommy Keane, a doctoral student in RIT’s Chester F. Carlson Center for Imaging Science from East Walpole, Mass., will present a paper he co-wrote with Cahill and Jeff Pelz, professor in the RIT Center for Imaging Science.

Keane won “Best Student Paper” at the workshop last year for a scientific technique he applied to extracting information from panoramic images. The paper he will present this year, “Image Sequence Event Detection via Recurrence Analysis,” continues his ongoing research in eye-tracking and computer vision, focusing on scene structure and understanding.

“I discuss in the paper a foundational example of how to search for significant eye-movement events, like fixations, saccades or smooth pursuit,” Keane says. “What this technique gives us is not only a way to visualize a potentially high-dimensional time-series, but a strong mathematical justification for detecting and quantifying statistical measures about how often, for how long, and in what way those eye-movement events occur throughout the video.”

Other RIT students who will present papers or posters include Gajendra Jung Katuwal, Thomas Kinsman, Victoria Scholl and Saugata Sinha from the Center for Imaging Science; Philip Hays, from computer engineering; and Kenny Davila and Vladimir Pribula from computer science.