Distance Transforms for Image Matching 
                         Daniel Huttenlocher
                         Cornell University

A distance transform of a binary image specifies for each 0-valued pixel the 
distance to the nearest 1-valued pixel (or vice versa). There are efficient 
algorithms for computing distance transforms in just a few operations per pixel 
making them of great practical use. Distance transforms form the basis for some 
of the most common binary image comparison techniques.  This talk will 
introduce distance transforms, provide some efficient algorithms for computing 
them, and cover their use in the Chamfer matching and Hausdorff distance 
techniques for binary image comparison.  It will also touch on more recent 
extensions of distance transforms to the kinds of combinatorial optimization 
problems that often arise in computer vision.

Dan Huttenlocher is the John P. and Rilla Neafsey Professor of Computing, 
Information Science and Business at Cornell University, where he holds a joint 
appointment in the Computer Science Department and the Johnson Graduate School 
of Management.  His research interests are in computer vision, geometric 
algorithms, interactive document systems, financial trading technology, and IT 
strategy.  Huttenlocher has 24 U.S. patents, has published more than 75 
technical papers, and has been recognized on several occasions for his teaching 
and research, including being named an NSF Presidential Young Investigator in 
1990, the New York State Professor of the Year in 1993, and a Stephen H. Weiss 
Fellow in 1996.  At Cornell, Huttenlocher chaired the Provost's task force that 
led to the creation of the new Faculty of Computing and Information Science.  In 
addition to his academic work, Huttenlocher has served as CTO of Intelligent 
Markets and was on the senior management team at Xerox PARC.