2008 WESTERN NEW YORK IMAGE PROCESSING WORKSHOP

Friday, September 26, 2008
Rochester Institute of Technology
Chester F. Carlson Center for Imaging Science Auditorium
Building 76
79 Lomb Memorial Drive
Rochester, NY 14623
(Parking Lots G and H)

Sponsored by the Rochester Chapter of the IEEE Signal Processing Society
and

Workshop Scope

The aim of this workshop is to promote interaction between image processing researchers in and around Western New York. Researchers from both academia and industry are invited to participate. Interaction will be encouraged by keeping the number of participants to less than one hundred, having a single track of presentations, and providing an informal lunch.

Plenary Talks

In addition to presentations of invited and solicited papers, the workshop will have two plenary talks delivered by renowned experts in their fields: Prof. Edward Dougherty, Texas A&M University and Dr. Ed Ashton, VirtualScopics, Inc. See below for titles, abstracts and bios.

Workshop Registration

Fees are kept to a minimum and are primarily intended to cover the costs of coffee breaks, lunch, and book of paper summaries.

IEEE Members: $20

Non-IEEE Members: $40

IEEE Member Students: $10

Non-IEEE Member Students: $20

Advanced registration is strongly encouraged by a substantial discount in the registration fees because we need a headcount for ordering the food. Hence, there will be a flat late fee of $10 for on-site registration. For advance registration, simply send an email to: vishalmonga@gmail.com by September 23rd. Please indicate any dietary restrictions and whether you are a student and an IEEE member.

All payments are due on-site by either check (payable to IEEE Rochester Section) or cash.

In order to encourage participation by students, we will present a best student paper award.

Workshop website: http://ewh.ieee.org/r1/rochester/sp/IP_workshop2008/workshop08.htm

Organizing Committee:

Chair: David Coumou, MKS Instruments, Inc.

John Handley, Xerox Corporation

Vishal Monga, Xerox Corporation

Andrew Gallagher, Eastman Kodak

Final Program

8:00AM~8:30AM	Registration and Welcome
8:30AM~9:30AM	Morning Plenary Statistical Design of Nonlinear Image Filters Edward R. Dougherty Department of Electrical and Computer Engineering, Texas A&M University and Computational Biology Division, Translational Genomics Research Institute
9:30AM~10:30AM	Session I: Image Security Exploiting  Spatial  Frequency  Separability for  Clustered-Dot Color Halftone Watermarking Basak Oztan and Gaurav Sharma University of Rochester, Rochester, NY Adaptive Decoding For Halftone Orientation Based Data Hiding Orhan Bulan and Gaurav Sharma ECE Dept, University of Rochester, Rochester, NY Vishal Monga Xerox Research Center Webster, Webster, NY Printer Characterization for UV Encryption Applications Yonghui Zhao and Raja Bala Xerox Research Center Webster, Webster, NY
10:30AM~11:00AM	Coffee Break

11:00AM~11:40AM	Session II: Image Security
	A Survey of Copy-Move Forgery Detection Techniques Sevinc Bayram Electrical and Computer Engineering Department, Polytechnic Institute of NYU, Brooklyn, NY Husrev Taha Sencar Computer Engineering Department, TOBB University of Economics and Technology, Ankara, Turkey Nasir Memon Computer and Information Science Department, Polytechnic Institute of NYU, Brooklyn, NY Video CAPTCHAs: Usability vs. Security Kurt Alfred Kluever and Richard Zanibbi Department of Computer Science, Rochester Institute of Technology, Rochester, NY
11:40AM~1:15PM	Lunch
1:30PM~2:30PM	Afternoon Plenary
	Quantitative Structural and Functional Medical Imaging Edward Ashton Chief Scientific Officer, VirtualScopics, Inc.
2:30PM~3:30PM	Session III: Image Understanding
	Inferring Generic Activities and Events from Image Content and Bags of Geo-tags Dhiraj Joshi and Jiebo Luo Eastman Kodak, Rochester, NY
	The Effect of Image Compression on Linear Kernel SVM Classifiers Grigorios Tsagkatakis Center for Imaging Science, Rochester Institute of Technology, Rochester, NY Andreas Savakis Computer Engineering, Rochester Institute of Technology, Rochester, NY Ordering Random Object Poses James Massaro and Raghuveer Rao Rochester Institute of Technology, Rochester, NY
3:30PM~3:50PM	Coffee Break
3:50PM~4:55 PM	Session IV: Color and Image Quality
	Finding Image Gamuts Using Expanding Sphere Techniques Marty Maltz Xerox Research Center Webster, Webster, NY N-color Moire-Free Halftoning Shen-Ge Wang and Robert Loce Xerox Research Center Webster, Webster, NY How to Use and Misuse Image Assessment Algorithms David M. Rouse and Sheila S. Hemami School of Electrical and Computer Engineering Cornell University, Ithaca, NY
4:55PM~5:00PM	Best Student Paper Awards Announcement

Statistical Design of Nonlinear Image Filters

Edward R. Dougherty

Department of Electrical and Computer Engineering, Texas A&M University

Computational Biology Division, Translational Genomics Research Institute

Up until around 1990, design of nonlinear filters for image processing was almost entirely ad hoc, the exception being in the case of some very simple and unrealistic models. This meant that nonlinear filtering was restricted to using a very small number of humanly designed structuring elements for morphological processing, along with simple order-statistic and stack filters. The situation was entirely different for linear filters, where the classical Wiener theory facilitated the design of optimal linear filters for many useful models. During the 1990s the theory and application of statistically designed optimal nonlinear filters dramatically changed the situation. Using image models and the theory of optimization, it became possible to construct complex filters for realistic image models and then to develop special architectures to implement filters consisting of thousands of structuring elements. But there was a price. Owing to nonlinearity, filter design swiftly ran up against computational limitations and had to address the complexity conflict inherent in pattern recognition: we desire high-complexity filters to more accurately recognize fine detail, such as that represented by high-frequency image structure; on the other hand, we desire low-complexity filters so that the designed filters do not overfit the training data, for instance, by conforming to high-frequency noise. This talk will briefly review the classical operators and then discuss optimal nonlinear filter design for both binary and gray-scale images in the context of general principles of pattern recognition.

Edward R. Dougherty is a Professor in the Department of Electrical and Computer Engineering at Texas A&M University in College Station, TX, where he holds the Robert M. Kennedy ‘26 Chair in Electrical Engineering and is Director of the Genomic Signal Processing Laboratory. He is also the Director of the Computational Biology Division of the Translational Genomics Research Institute in Phoenix, AZ. He holds a Ph.D. in mathematics from Rutgers University and an M.S. in Computer Science from Stevens Institute of Technology, and has been awarded the Doctor Honoris Causa by the Tampere University of Technology in Finland. He is a fellow of SPIE, has received the SPIE President’s Award, and served as the editor of the SPIE/IS&T Journal of Electronic Imaging. At Texas A&M he has received the Association of Former Students Distinguished Achievement Award in Research, been named Fellow of the Texas Engineering Experiment Station, and named Halliburton Professor of the Dwight Look College of Engineering. Prof. Dougherty is author of fourteen books, editor of five others, and author of more than two hundred journal papers. He has contributed extensively to the statistical design of nonlinear operators for image processing and the consequent application of pattern recognition theory to nonlinear image processing. In recent years he has helped lead the development of genomic signal processing, which is aimed at diagnosis and prognosis based on genetic signatures and uses gene regulatory networks to develop therapies based on the disruption or mitigation of aberrant gene function contributing to the pathology of a disease.

Quantitative Structural and Functional Medical Imaging

Edward Ashton

Chief Scientific Officer, VirtualScopics, Inc.

The use of medical imaging for diagnostic purposes as well as for evaluation of pharmaceutical trials has increased exponentially in recent years. Cross-sectional imaging techniques such as MRI and CT allow views of anatomical structures that are in many cases superior to those obtainable through exploratory surgery, while molecular and functional imaging allow the direct observation of oxygen metabolism, receptor binding, and blood flow in vivo. However, evaluation of these images is still largely qualitative. This talk will examine image and signal processing techniques that are currently being brought to bear to allow quantitative analysis of biological parameters from medical images in clinical trials, with an eye toward future applications in the diagnostic arena.

Edward Ashton serves as Chief Scientific Officer for VirtualScopics, Inc., where for the past eight years he has had primary technical responsibility for all projects in both oncology and central nervous system disease. He has extensive experience in image acquisition and analysis in both biomedical imaging and military surveillance and reconnaissance. Prior to joining VirtualScopics, Dr. Ashton was a lead signal processing engineer at The MITRE Corporation in McLean, VA. Earlier in his career, he served as a research engineer with the Naval Research Laboratory, where he received the Alan Berman Research Publication Award and was nominated for the Edison Award for Applied Science. Dr. Ashton has produced numerous articles on biomedical imaging as well as target detection and image analysis with military applications. He received both his Ph.D. and M.S. degrees in electrical engineering from the University of Rochester, where his research focused on MRI applications in the brain, and his B.S. degree in electrical engineering from Loyola College.