Theoretical Foundations of Computer Vision -- Geometry, Morphology, and Computational Imaging

Motivation

Image analysis and computer graphics depends on geometric modelling and analysis of objects in two- or multidimensional spaces. Different disciplines such as digital geometry, mathematical morphology, polyeder geometry or computational geometry, just to cite a few, are closely related to progress in image analysis and computer graphics.

The seminar intention is to discuss theoretical fundamentals related to those issues and to specify open problems and major directions of further development in the field of geometric problems related to image analysis and computer graphics. The seminar schedule will be characterised by flexibility, working groups, and sufficient time for focused discussions. There will be an edited volume of seminar papers (within the Springer LNCS series) with a submission date following the workshop.

The co-chairs propose to consider contributions related to one of the following subjects:

• geometric algorithms for computer vision for extracting structures from images, geometric shape matching, image segmentation and image restoration,
• mathematical morphology (shape descriptors, connectivity models, topological algorithms, grey-level and colour image geometry),
• geometric feature analysis (length of a curve, area of surfaces in 3D, curvature: models, algorithms, performance analysis, multigrid convergence),

All seminar participants are expected to contribute either a short (20 min) or a long (40 min) talk during the seminar. Some contributions in neighbouring topics (digital topology, arithmetical geometry, stochastic geometry, geometric modelling) can also be accepted, however, the majority of talks should focus on the specified subject list.

Participants may contribute to a list of open problems for inclusion into the workshop report. Working groups may be formed on day one of the seminar for smaller and more focuses meetings during the week. There will be a session on Monday for discussing options for such working groups.

Reports of the previous Seminar on the Theoretical Foundations of Computer Vision are still available at Dagstuhl office or via the Dagstuhl webserver. The recent ones are: