DISI
Universita di Genova - Dipartimento di Informatica e Scienze dell'Informazione (Italy)
The Geometric Modeling and Computer Graphics group,
funded in 1990, is active in the fields of
geometric and solid modeling, computer graphics and computational geometry. The current research activity is focused on modeling and computational issues related to the representation and manipulation of shapes and scalar fields. The expertise in the CV group includes motion analysis, interpretation of 3D scenes,
object and pattern recognition, machine learning. The entire DISI team has been working on applications of shape representation and recognition techniques in Geographic Information Systems (GISs), in augmented virtual reality, in cultural heritage and biomedicine.
The research activity on shape representation and manipulation concerns the following topics
- Multi-resolution representation of multi-dimensional scalar fields for applications to scientific data visualization and implicit surface modeling: in particular, our research focuses on Level-Of-Detail (LOD) models based on nested grids for regular volume data sets and time-varying scalar fields, and on LOD models for unstructured volume data sets.
- Multi-scale models for non-manifold shapes: in particular, our research focuses data structures and algorithms for encoding and manipulating non-manifold shapes described through simplicial meshes, simplification of non-manifold shapes, part-based representations.
- Morphology-based representations of terrains (2D scalar fields) and volume data sets
-
Surface reconstruction for virtual reality applications: our research focuses on on-line mosaicing to reconstruct 3D objects from multiple acoustic/optical range images, and on techniques reconstructing surfaces from data with uncertainties coming also from multiple sensors.
Active projects on shape representation and manipulation are
- EC project ARROV (http://arrov.disi.unige.it)
- Basic research project funded by the Ministry of Education, University and Research (MIUR) on "Algorithmic and Computational Methods for Geometric Object Representation (MACROGeo)" (with IMATI)- 2003-2004
- MIUR basic research project on "Representation and management of spatial data on the WEB (WEB-GIS)"-2003-2004
-
Scientific Research Programs of
Relevant Narional Interest - Ministry of Education, University and Research
(PRIN-MIUR), Projects N.2003011918_006,
"Low-cost sensors and geometric techniques for scanning and modeling solid objects" / 2003-2005
The Computer Vision group
performs research on the basic computational problems underlying the analysis and
the understanding of digital images. The aim is designing, developing, and testing image processing methods for:
- extracting 3-D geometrical structure from stereo, active triangulation, or image sequences;
- dealing with several computer vision problems like image filtering, feature extraction and grouping, camera calibration, and motion analysis.
Further information can be found at the following URL: