Softwares
Name: Co-Registration of Satellite Imagery
Description: CO-REGIS is a satellite image processor for the co-registration of Sentinel-2 and Landsat-8 imagery and the accurate correction of co-registration residuals allowing time series analysis.
Input data: Copernicus Sentinel-2, NASA Landsat-8
Output products: Co-registered image time series
Application: Creation of datacube (image and derived features)
License: 
References: Stumpf et al. (2018)
Contact: Jean-Philippe
Name: Data cube for Sentinel-2 imagery
Description: DataCube is a satellite image processor for generating consistent stacks of spatially aligned pixels, and derived features, for time-series, allowing easy access to Analysis Ready Data (ARD).
Input data: Copernicus Sentinel-2
Output products: Stacks of consistent aligned pixels with chosen high-level variables (reflectance, spectral indices, …)
Application: Detection and quantification of landcover and landuse change trajectories
License:
References: Stumpf et al. (2018)
Contact: David
Name: Digital Surface Model from Optical Imagery
Description: DSM-OPT is a satellite image processor for the generation of Digital Surface Models of the Earth surface from stereo-pairs of satellite imagery. It is based on the MicMac open source library.
Input data: Pléiades stereo and tri-stereo pairs
Output products: Digital Surface Models representing the elevation of the natural or built topography
Application: Creation of relief maps
License:
References: Stumpf et al. (2019)
Contact: Jean-Philippe
Name: URBan Footprint from OPTical Imagery
Description: URBA-OPT is a satellite image processor for urban footprint detection. It is based on a supervised object-oriented approach using machine learning (Random Forest).
Input data: Copernicus Sentinel-2
Output products: Urban footprint maps and associated accuracy
Application: Inventory and mapping of artificialized areas
License:
References: Puissant et al. (2018)
Contact: Anne
Name: Detection of Surface WATER from Sentinel-2
Description: Water-S2 is a satellite image processor for the detection and monitoring of surface waters from Sentinel-2 imagery. It is based on a supervised approach using machine learning (Random Forest).
Input data: Copernicus Sentinel-2
Output products: Surface water maps and water frequency maps
Application: Permanent and temporary (flooding) surface water mapping
License:
References: Faivre et al. (2018)
Contact: Bernard
Name: Detection of Surface WATER from Sentinel-1
Description: Water-S1 is a satellite image processor for the detection and monitoring of surface waters from Sentinel-1 imagery. It is based on finite mixture modeling and bilateral filtering for detecting the surface waters.
Input data: Copernicus Sentinel-1
Output products: Surface water maps and water frequency maps
Application: Permanent and temporary (flooding) surface water mapping
License:
References: Bioresita et al. (2018) | Bioresita et al. (2019)
Contact: Anne
Name: Multiple Pairwise Image Correlation
Description: MPIC-OPT is a satellite image processor for the analysis of optical satellite image time-series for the monitoring of Earth surface deformation (earthquake, landslides, ice glaciers).
Input data: Copernicus Sentinel-2, NASA Landsat-8, Spot 6/7, Pléiades
Output products: Surface deformation maps and associated accuracy
Application: Quantification of the motion of landslides, Quantification of tectonic movements
License:
References: Stumpf et al. (2014) | Stumpf et al. (2018)
Contact: Jean-Philippe
Name: User-tailored IMage CLASSification for land surface mapping
Description: imCLASS is a satellite image processor for the detection of object of interest (presence/absence) of the land surface. It is based on a supervised change detection method using machine learning (Random Forest) and active optimization of the training samples.
Input data: Copernicus Sentinel-2, NASA Landsat-8, Spot 6/7, Pléiades
Output products: Inventory of object of interest
Application: Detection and mapping of any objects of interest over large areas with only a need of a few training samples
License:
References: Deprez et al. (2018)
Contact: Aline
Name: Detection of fire BURNT scars
Description: BURNOUT is a satellite image processor for the detection of land surfaces affected by wildfires from Sentinel-2 imagery. It is based on change detection techniques and thresholding of image features for detecting fire burnt scars.
Input data: Copernicus Sentinel-2, NASA Landsat-8
Output products: Burnt area maps
Application: Detection of fire burnt scars and severity estimation
License:
References: Caspard et al. (2018)
Contact: Mathilde
Name: A2S-Data - A2S data discovery
Description: This code retrieves satellite data from several repositories and associates two functions: a meta-request function consisting in emitting requests to interogate all the registered repository catalogs, and a meta-download function consisting in downloading the products from the easiest reachable provider at the time of the request; in case of download problems, the code cascades to other data providers.
License:
Contact: David
Name: A2S-WMS - A2S Workflow Management System
Description: This code consists in a Workflow Management System (e.g. based on Fireworks WMS) used to express task parallelism inside a single calculation job and among several calculation jobs. It provides a monitoring interface along with a powerfull CLI to allow control, correction and restart of the tasks in case of problems during a run. The code is developped in order to not lose any job.
License:
Contact: David
Name: A2S-Scheduler - A2S task scheduler (built on top of A2S-WMS)
Description: This code allows submitting several tasks in an efficient manner for different type of calculation jobs, either on all the cluster nodes, sockets or cores. It allows optimizing the computing ressource usage. It is able to interrogate the WMS database and the cluster workload manager (e.g.in our case, Slurm) knowing at the same time the global ressources needed for the jobs, the local ressources available on the node (free sockets/cores) and the global ressources available on the cluster. Using a Master/Slave control pattern, the instances of the task scheduler are able to adapt the provisionning of nodes to the computational needs (submitting or releasing nodes), following the computational workload.
License:
Contact: David
Name: A2S-Cache - A2S cache management system
Description: This code allows reducing data movement by permitting to write the data, during computation, on a fast and close File system (e.g. cache system) being at the same time duplicated and pushed on a storage server. The cache is periodically cleaned from all files older than a given period. In case of cache miss, the cache can be rebuilt on fly transparently.
License:
Contact: David
