GEOSTOSIP · Geometrical Stochastic Signal Processing: Applications to wave Physics in random media
FP7 — People (Marie Curie Actions)
- Duration
- 2013-08-01 → 2016-07-31
- EU contribution
- €230,937
- Participants
- 1
- Scheme
- MC-IOF
Lines connect the coordinator with its partners.
Results in brief
Geometrical Stochastic Signal Processing: Applications to wave Physics in random media
The Geostosip project has led to the publication of results on new algorithms for the signal processing community. The definition of new multiple scattering processes has been proposed, together with new estimation algorithms dedicated to the study of these processes. The use of geometric harmonic analysis has allowed the characterization of the geometric phase for multivariate signals. The case of bivariate signals has been particularly investigated. The results of the Geostosip have potential applications in research fields where researchers face vector motions in planes or spaces. The new geometric signal processing paradigm described in the publications linked with the Geostosip research project should bring a new way of dealing with multivariate signals and signals taking values on manifolds. In particular, research areas including seismology and ultra-sound imaging should benefit from the algorithms proposed in the framework of Geostosip.
Data: CORDIS, © European Union
Project objective
The research project deals with the geometry of random and disordered media at a mesoscopic level and its study from a signal processing perspective. The main application is wave physics in random and complex media.Two topical and innovative questions are addressed.1- Can we obtain a multiscale geometrical image of a random/complex medium from observations of waves properties at the output of the medium, using statistical measures ?2- Can we observe the dynamics of a random medium from geometric properties of the output signals and its geometrical spectrum ?The goal is to develop a complete methodology to characterize random media using probing waves. The major innovation proposed in the project is to make use of random processes on manifolds to model the behaviour of waves in random media and to develop some geometrical methods (differential geometry, probability and statistics on manifolds and Geometric/Clifford harmonic analysis) allowing the extraction of structural informations about the media.
Original text from CORDIS.
Participants
- CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE CNRS · ParisCoordinatorFrance
Links
Data: CORDIS, © European Union