MultiSenGE-Tools

This repository contains few tools to sort and extract stats on MultiSenGE dataset. If you want to download the dataset, please follow this link

If you use this script and/or MultiSenGE dataset, please cite the paper as follow.

Romain Wenger, Anne Puissant, Jonathan Weber, Lhassane Idoumghar, & Germain Forestier. (2022). A new remote sensing benchmark dataset for machine learning applications : MultiSenGE (1.0) [Data set]. Zenodo. https://doi.org/10.5281/zenodo.6375466

@dataset{romain_wenger_2022_6375466,
  author       = {Romain Wenger and
                  Anne Puissant and
                  Jonathan Weber and
                  Lhassane Idoumghar and
                  Germain Forestier},
  title        = {{A new remote sensing benchmark dataset for machine 
                   learning applications : MultiSenGE}},
  month        = mar,
  year         = 2022,
  note         = {ANR-17-CE23-0015},
  publisher    = {Zenodo},
  version      = {1.0},
  doi          = {10.5281/zenodo.6375466},
  url          = {https://doi.org/10.5281/zenodo.6375466}
}

And/or

Wenger, R., Puissant, A., Weber, J., Idoumghar, L., and Forestier, G.: MULTISENGE: A MULTIMODAL AND MULTITEMPORAL BENCHMARK DATASET FOR LAND USE/LAND COVER REMOTE SENSING APPLICATIONS, ISPRS Ann. Photogramm. Remote Sens. Spatial Inf. Sci., V-3-2022, 635–640, https://doi.org/10.5194/isprs-annals-V-3-2022-635-2022, 2022.

@Article{isprs-annals-V-3-2022-635-2022,
  AUTHOR = {Wenger, R. and Puissant, A. and Weber, J. and Idoumghar, L. and Forestier, G.},
  TITLE = {MULTISENGE: A MULTIMODAL AND MULTITEMPORAL BENCHMARK DATASET FOR LAND USE/LAND COVER REMOTE SENSING APPLICATIONS},
  JOURNAL = {ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences},
  VOLUME = {V-3-2022},
  YEAR = {2022},
  PAGES = {635--640},
  URL = {https://www.isprs-ann-photogramm-remote-sens-spatial-inf-sci.net/V-3-2022/635/2022/},
  DOI = {10.5194/isprs-annals-V-3-2022-635-2022}
}

Prerequisites

Normally, json is a built-in library available in Python. In case it is not installed, I invite you to set it up with conda :

conda install -c jmcmurray json

List of methods in Patch class

Methods :

• reconstruct_filename (Public) : Reconstruct the filename(s) of ground reference (gr) patch, Sentinel-2 (s2) patche(s) or Sentinel-1 (s1) patches.

• matching_periods (Public) : Check if the patch have nb_data_per_period for each period in periods (list of a list)

• has_nb_dates (Public) : Check if the patch has a certain number (nb_data patchs) of dates.

• has_matching_monthes (Public) : Check if the patch has a certain number (nb_data_per_month patchs) for each months (dates) for S1 or S2.

• get_centroid (Public) : Calculating the centroid of the patch.

• create_points_shapefile (Public Static) : Compute shapefile map from a list of points/centroids (default EPSG : 4326).

• change_coordinates (Public Static) : Changing coordinates from an epsg to an other.

• has_days_gap_s2 (Public) : Check if there is at least days_gap between each S2 image in each month.

• to_date (Public Static) : Convert a list of string dates in a format (date_format). Possibility to extract a date from a filename (S1 or S2 patches filename)

• generate_list_patches (Public Static) : Create a list of patches objects from json files available in MultiSenGE (labels folder).

Examples :

• This function read every json file in the folder. It returns a list containing x Patch objects according to the number of patches present in the dataset.

list_patches = Patch.generate_list_patches('./labels')

• There are two ways to use the constructor of the class. The first one is to give it a json filename from the labels folder.

constructor_wt_filename = Patch(json=os.path.join('./labels', '31UGP_2313_4883.json'))

The second one uses kwargs and you have to instantiate it with every parameters.

constructor_wt_params = Patch(x=x, y=y, tile=tile, s2_dates=s2_dates, s1_dates=s1_dates, projection=projection, labels=labels)

• With this function, you want to know if the current patch (mypatch) has at least 2 S2 patches for each period in periods.

periods = [['20200101', '20200731'], ['20200801', '20200930'], ['20201001', '20201231']]
mypatch.matching_periods(periods, 2)

Dataset visualization :

You can follow this link to visualize some informations extracted from MultiSenGE.

Contact

This script was made by Romain Wenger. For any questions or suggestions, do not hesitate to contact us : Mail me