7/13/2009 Susan Kantor, ECE ILLINOIS
ECE Associate Professor Minh Do recently received the Young Author Best Paper Award from the IEEE Signal Processing Society. His paper describes computational techniques for visual information representation that mimic the way that the human visual system works.
Written by Susan Kantor, ECE ILLINOIS
ECE Associate Professor Minh N Do recently received the Young Author Best Paper Award from the IEEE Signal Processing Society. His paper, “The Contourlet Transform: An Efficient Directional Multiresolution Image Representation,” co-authored by Martin Vetterli, describes computational techniques for visual information representation that mimic the way that the human visual system works.
“The question of how to come up with a very efficient way of representing visual information is the fundamental problem in image processing and many other visual information disciplines,” Do said.
When visual information is represented more efficiently, its transfer becomes more efficient as well. This means more images can be stored on a digital camera, more television channels can be transferred through a single cable and images can be downloaded faster on the Internet.
The Fourier transform, which has been in existence for nearly 200 years, is used to store and compress images into the JPEG format. Within the last 20 years, the wavelet transform has become an image-compression standard. JPEG 2000 is what is typically seen in digital movie theaters. Do’s research leads to the contourlet transform, which he hopes will be in the next standard for image compression.
Do began this research during his final year as a PhD student at the Swiss Federal Institute of Technology Lausanne, Switzerland in 2000. The paper was submitted in 2003 and published in 2005.
“I got known because of this work,” Do said. “I think the main reason I got hired by Illinois is really this work.”
When Do came to Illinois, he wanted to expand the research. How can the system be better designed? How can it be fine tuned? How can it be generalized? How can these ideas be expanded to video?
“Many of those next-step questions were then laid out and addressed by my PhD students,” Do said.
Do’s paper has been cited in more than 500 other papers, which is a reason why Do’s paper won the award. The techniques have been applied for other applications, such as analyzing medical images and enhancing image quality.
Another reason why it was chosen was because there is an accompanying software package publicly available. There are more than 100 downloads of the software per month, Do said.
“This is a paper that is very dear to my heart,” Do said. “It was a springboard for some of my first PhD students. It put our group on the map. After a few years at Illinois, our group has become one of the top groups in this area of so-called image representation and wavelets. When the paper was formally recognized, certainly we were very happy about that.”
Do has been at the University since 2002. He is a Senior Member of IEEE and a Beckman Fellow at the Center for Advanced Study. He has also received the Xerox Award for Faculty Research from the College of Engineering in 2007, the CAREER award from the National Science Foundation in 2003 and Best Doctoral Thesis Award from the Swiss Federal Institute of Technology Lausanne in 2001.