Length, width, and relative age analysis of lineaments in the Galileo regional maps with LineaMapper (Haslebacher et al., PSJ, 2025) accompanying data

Published: 11 June 2025| Version 2 | DOI: 10.17632/rjhsjrnxgv.2
Contributors:
,
,
,
,
,

Description

Data accompanying publication "Length, width, and relative age analysis of lineaments in the Galileo regional maps with LineaMapper", published in the Planetary Science Journal (PSJ), in the special issue "Origins and Habitability of the Galilean Moons". The model weights are used in the google colab jupyter notebook linked on github: https://github.com/javirk/europa_surface This data repository includes: - the unrevised lineament map (output of LineaMapper v1.0, in 'Shapefiles_17ESREGMAP02/LM1_0_initial_prediction') - revised lineament map (full map of 17ESREGMAP02 ('Shapefiles_17ESREGMAP02'); automatically analysed and manually verified chaos and ridged plains regions (in 'analysed_CSVs_17ESREGMAP02')) - the geotiff files used for mapping (in 'Geotiffs_17ESREGMAP02', based on Bland et al., 2021) - the geotiff files for predictions (you only need the .cub files to run apply_regmaps_improved_LineaMapper_v2_to_all_galileo.py, but geotiffs are also provided for convenience) - full LineaMapper v1.1 and v2.0 regional maps (in 'full_RegMaps_shapefiles' for quick download, and additional files in zips 'analyse_LM1_1.zip' and 'analyse_LM2_0.zip') - train, validation, and test tiles for LineaMapper v1.0, v1.1 and 2.0 (in 'datasets_train_val_test') - model weights (folder 'ckpts', LM1.0: Mask_R-CNN_pub2_run23_end_model.pt, LM1.1: Mask_R-CNN_v1_1_17ESREGMAP02_part01_run10_end_model.pt, LM2.0: bbox_vit_b_final.pt, other files are other versions of LM2.0 (see paper)) - code (isis, LM1.1, LM2.0 from github) To get started with using LineaMapper, the github repository provides jupyter notebooks that can be run in a web browser through Google Colab. The notebooks access this Mendeley dataset to download the model weights. The 'code' folder also contains these jupyter notebooks and they can be run locally as well.

Files

Institutions

NASA Jet Propulsion Laboratory, Southwest Research Institute Boulder, Johns Hopkins University Applied Physics Laboratory, Universitat Bern Philosophisch-naturwissenschaftliche Fakultat

Categories

Planetary Science, Image Segmentation, Deep Learning, Instance Segmentation

Funding

Swiss National Science Foundation

P500PT_225447

National Centre of Competence in Research PlanetS

51NF40_205606

National Centre of Competence in Research PlanetS

51NF40_182901

Licence