Vous êtes ici : GIPSA-lab > Animation > Séminaires
Chargement

Séminaire du département Images et Signal du 20/02/2020 à 14h00

 

A graph Laplacian-based multimodal feature selection method for efficient remote sensing data analysis

Intervenant : Saloua CHLAILY, UiT - The Arctic University of Norway

Lieu : Salle Chartreuse

 

Résumé :

When dealing with multivariate remotely sensed records, collected by multiple sensors, an accurate selection of information at data, feature, or decision level is instrumental in improving the characterization of the scenes, enhancing the efficiency of the system, and providing more details on modeling the physical phenomena occurring on Earth's surface. In my talk, I will introduce a flexible and efficient method, based on Graph Laplacians, for information selection at different levels of data fusion. The selection is performed patch-wise, as such different images, features, or decisions will be chosen according to the properties of each homogeneous area. Experimental tests carried out on several instances of multivariate remote sensing datasets show the consistency of the proposed approach.


GIPSA-lab, 11 rue des Mathématiques, Grenoble Campus BP46, F-38402 SAINT MARTIN D'HERES CEDEX - 33 (0)4 76 82 71 31