Post-doc position in data mining and remote sensing

2/28627
label Burse autorenew 2025-09-29, 17:00 history_edu Ioana Dinescu
Subject: Exploiting multi-resolution remote sensing images for change detection and classification purposes.
LIRMM and TETIS (in collaboration with Gloria Bordogna CNR IDPA - Italia) are seeking candidates for a Postdoctoral Fellowship position to work on a new research project, called GEOSUD, to exploit multi-resolution remote sensing images for land cover change detection and classification purposes.

More in detail the aim of the work is to improve the quality of dense time series of medium or low resolution images by means of rare high resolution time series. The satellite images are acquired over well documented study sites concerned with land and vegetation management issues. The first site is located in Reunion Island (a French territory in the Indian Ocean) and is concerned by the sugarcane crop monitoring using multi-resolution œSPOT-Pleiades combinations of images. The second one is situated in West Africa and is concerned by land use mapping using multi-resolution MODIS-SPOT combinations of images.



From a computer science point of view, we intend to apply soft computing techniques and fuzzy aggregation operators in order to flexibly fuse and propagate the high resolution information (as satellite images) in the low resolution one. The goal is to obtain a dense time series of better quality w.r.t. the original one. The final goal of the application of soft constraints aims to improve identification of particular types of crop, and monitoring the crop conditions.

Experiments will be performed at different levels (pixel, feature and decision) by associating different weights to the corresponding value to fuse in order to take into account the correlation between the two types of original images and the time interval of their acquisitions.