Presents a sampling of the research issues related to applying graph-based methods in computer vision. These methods have been under-utilized in the past, but use must now be increased because of their ability to naturally and effectively represent image models and data. This publication explores current activity and future applications of this fascinating and ground-breaking topic.
