IMAGeS team: IMages, leArning, Geometry and Statistics

Loïc Mazo

From IMAGeS team: IMages, leArning, Geometry and Statistics
Revision as of 09:14, 3 April 2013 by Mazo (talk | contribs)
Jump to navigation Jump to search

Assistant Professor

ICube - MIV
300 Bd Sébastien Brant
BP 10413
67412 Illkirch CEDEX - France

Office : C217

Phone : + 33 (0) 3 68 85 44 96
( IUT : + 33 (0) 3 68 85 84 31 )
E-mail : loic.mazo@unistra.fr

Mazo.jpg

Description of Activities

My activities fall within the mathematics of discrete imagery: topology, geometry, measure.

Discrete topology

I am interested in topological models for binary discrete images per se, that is, that do not depend on some underlying Euclidean reality. This include the graph-theoretical approach or the cellular approach and does not exclude appealing to notions built on continuity, as homotopy, thanks to the use of partial orders on domains. Besides, I am working on extending such models to images whose codomains are partially ordered sets -- from labels images to grey-level images -- so we can take into account not only the topology of each object in an image but also the topological relations between these objects.

Discretization

My works focus on Hausdorff distance based discretizations with the goal to extend such discretization schemes to grey-level and color images while describing their geometrical, topological and convergence properties.

Estimation of geometrical parameters

I study how to start the construction of a measure and integration theory on a discrete space that must have some convergence property towards the equivalent theory in the Euclidean space.

Tomography

My research is concerned with deformable object reconctruction from non-oriented projections.

Publications

<anyweb> http://newlsiit.u-strasbg.fr/papr/appli.php?author=mazo&title=&labo=tous&team=toutes&annee1=&annee2=&display=rap+&nationalRank=toutes&project=tous&hide=0&hide=0 </anyweb>