Abstract

Air project is involved in earth, sea and atmospheric dynamic image sequences acquired with remote sensors. Air project is prospecting in three directions: image processing problems, thematic applications and analysis of data acquired by different remote (satellite) sensors. Each research activity in the project may be considered in view of the latter three directions.

From a scientific point of view, AIR project's goal is to set up image processing capabilities for dynamic remotely sensed data. Research activities are grouped into two main categories:

1.: Structures' modeling for remotely sensed data of earth-sea-atmospheric observation. This topic encompasses modeling and tracking methods for prediction about clouds, vortex, landslides...
2.: Study of the temporal evolution of natural phenomena, which involves motion, deformation and coupled SPOT-NOAA subpixel analysis.

The project is interested in various thematic applications ranging from meteorology, forest and vegetation evolution to oceanography and landslides; such studies are run in collaboration with scientists working in these areas. They submit the problems and validate both results and problems accuracy.

Lastly, specific attention is dedicated to intrinsic properties of images obtained with different sensors: SPOT, NOAA, METEOSAT... and to the ability of mutual use for multisensors models with variable temporal and spatial resolutions.

The objective of AIR project over the long term is participating in the creation of an image processing and managing system of satellite images for environmental applications involving evolution monitoring and alert triggering.

Précédent : Références Remonter : Projet AIR, Traitement d'image et