LandCoM: Land cover mapping

Description

The LandCoM project aims to develop an automated method for evaluating by remote sensing the percentage of soil cover by crop residues and cover crops.

Objective(s)

Develop and validate a method for analyzing satellite remote sensing images to distinguish and quantify in the early spring period the percentage of soil cover by grassland, cover crops and crop residues on a return of annual field crops.

Carried out in collaboration with INRS-ETE, the analytical approach of CarTéCoS is based on a three-scale methodology:

1. micro-scale: the collection and analysis of digital photos taken using smart phones;
2. mesoscale: the collection and analysis of high-resolution spatial multispectral images obtained by a drone;
3. macro-scale: the analysis of multi-source and multi-temporal Sentinel 2 and Landsat-8 (optical) and Sentinel 1 radar images by automatic deep learning algorithms in the Google Earth Engine © cloud platform.