SDC, Data Science and Knowledge

Difference between revisions of "Stella Marc-Zwecker"

From SDC, Data Science and Knowledge
Jump to navigation Jump to search
Line 17: Line 17:
  
  
* FOURIRE (Fuzzy Ontologies for URban Image REcognition) - API ICube (2014-2016) : this project, in collaboration with [http://icube-miv.unistra.fr/ MIV] team, involves the systemic consideration of vagueness and uncertainty in the whole process of urban object recognition. The consideration of uncertainty begins from the step of segmentation of the satellite image (fuzzy segmentation, "fuzzy ground truth"), and ends to the level of the urban object construction (by applying a fuzzy qualitative spatial reasoning).
+
* FOURIRE (Fuzzy Ontologies for URban Image REcognition) - API ICube (2014-2016) : this project, in collaboration with [http://icube-miv.unistra.fr/ MIV] team, involves the systemic consideration of vagueness and uncertainty in the whole process of urban object recognition. The consideration of uncertainty begins from the step of segmentation of the satellite image (fuzzy segmentation, "fuzzy ground truth"), and ends at the level of urban object construction (by applying a fuzzy qualitative spatial reasoning).

Revision as of 17:44, 4 March 2016

Photo stella2.JPG

Associate professor at the Computer Science Department of the University of Strasbourg.

Research areas

My research in knowledge representation is mainly applied to the field of semi-automatic recognition of high resolution satellite urban images :

  • Domain knowledge modeling using ontologies (semantic web).
  • Spatial and temporal qualitative reasoning.
  • Modeling vague and imprecise knowledge within ontologies using fuzzy logic.

Research Projects

  • COCLICO (COllaboration, CLassification, Incrementality and Knowledge) - ANR MN (2012-2016) : my contribution to this project mainly consists in the design of a software platform based on an "image" ontology, which integrates a fuzzy reasoning, that provides a fuzzy classification of image segments into primary classes (vegetation, shadow, water, soil, artificial. This fuzzy classification aims to help the process of segments grouping into urban objects of interest (buildings, roads, parks, etc.).


  • FOURIRE (Fuzzy Ontologies for URban Image REcognition) - API ICube (2014-2016) : this project, in collaboration with MIV team, involves the systemic consideration of vagueness and uncertainty in the whole process of urban object recognition. The consideration of uncertainty begins from the step of segmentation of the satellite image (fuzzy segmentation, "fuzzy ground truth"), and ends at the level of urban object construction (by applying a fuzzy qualitative spatial reasoning).