Geodesic Methods in Computer Vision and Graphics

Geodesic methods are now a standard tool in computer vision and graphics. They offer a unifying framework to describe the local geometry of images, surfaces and higher dimensional datasets. Fast and efficient algorithms compute geodesic distances to a set of points and shortest paths between points. This paves the way towards powerful methods to solve many important problems in vision and graphics such as image segmentation, surface meshing and shapes comparison. Geodesic Methods in Computer Vision and Graphics is an extended review of this emerging field and features the - The mathematical foundations underlying these methods are explained in a clear and theoretically sound way; - State of the art algorithms to compute shortest paths are thoroughly discussed; - Several fields of application are reviewed, including medical imaging segmentation, 3-D surface sampling and shape retrieval. This book is of interest to students and researchers in computer vision and graphics who wish to understand the foundations and the recent developments in this area.