Skip to content

feat(providers): add spacenet collections to dedl #1857

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
sbrunato merged 9 commits into from
Apr 13, 2026
Merged

feat(providers): add spacenet collections to dedl #1857

Show file tree
Hide file tree
Changes from 6 commits
Commits
Show all changes
9 commits
Select commit Hold shift + click to select a range
72ed74c
feat: spacenet
cauriol Sep 26, 2025
a934145
feat: add spacenet to dedl
Feb 18, 2026
e7e5db3
fix: tests
Feb 18, 2026
f170e0f
fix: precomit
cauriol Feb 18, 2026
5c6d45b
fix: collections parameters
cauriol Feb 19, 2026
d7450e7
fix: remove missionStartDate
cauriol Feb 20, 2026
05642bb
Update eodag/resources/collections.yml
cauriol Apr 10, 2026
9a6689b
Update eodag/resources/collections.yml
cauriol Apr 10, 2026
8cd2a49
Apply suggestions from code review
cauriol Apr 10, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
80 changes: 80 additions & 0 deletions eodag/resources/collections.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -7128,6 +7128,86 @@ MO_OCEANCOLOUR_GLO_BGC_L4_MY_009_108:
title: Global Ocean Colour Plankton MY L4 monthly observations
extent: {"spatial": {"bbox": [[-180.0, -90.0, 180.0, 90.0]]}, "temporal": {"interval": [["1997-09-01T00:00:00Z", "2024-04-01T00:00:00Z"]]}}

# MARK: spacenet ------------------------------------------------------------------------
SPACENET_BUILDINGS_DETECTION_V1:
Comment thread
jlahovnik marked this conversation as resolved.
description: |
The first SpaceNet challenge focused on large-scale building detection. It provided 2,544 km² of high-resolution WorldView satellite imagery over Rio de Janeiro, Brazil, along with 382,534 building footprint labels. Both 3-band and 8-band imagery were included to support algorithm development for automated building extraction.
instruments: []
constellation: SpaceNet
platform: SN1
processing:level:
keywords: ["Rio", "Building", "Building Detection v1", "Building Detection", "v1", "SpaceNet", "Chipped Training Dataset"]
eodag:sensor_type:
Comment thread
cauriol marked this conversation as resolved.
Outdated
license: CC-BY-4.0
title: "SpaceNet 1: Building Detection v1"
extent: {"spatial": {"bbox": [[-43.7753747662, -23.0106794336, -43.4847465904, -22.8256549561]]}, "temporal": {"interval": [["2016-02-29T00:00:00Z", null]]}}

SPACENET_BUILDINGS_DETECTION_V2:
description: |
The SpaceNet 2 dataset expands the building detection challenge to multiple cities worldwide, including Las Vegas, Paris, Shanghai, and Khartoum. It consists of high-resolution WorldView satellite imagery with annotations for more than 300,000 building footprints across 665 km². The dataset is designed to evaluate model generalization and robustness in detecting buildings across diverse urban landscapes.
instruments: []
constellation: SpaceNet
platform: SN2
processing:level:
keywords: ["Vegas", "Paris", "Shanghai", "Khartoum", "Building", "Building Detection v2", "Building Detection", "v2", "SpaceNet", "Chipped Training Dataset"]
eodag:sensor_type:
Comment thread
cauriol marked this conversation as resolved.
Outdated
license: CC-BY-4.0
title: "SpaceNet 2: Building Detection v2"
extent: {"spatial": {"bbox": [[-115.3075176, 15.5138111999, 121.7498742, 49.0582359 ], [-115.3075176, 36.1212776997, -115.1513226, 36.2616777], [2.18641139997, 48.9757509, 2.30048639992, 49.0582359], [121.5901692, 31.1978709001, 121.7498742, 31.4277759], [32.4858384, 15.5138111999, 32.5665684, 15.7402062]]}, "temporal": {"interval": [["2015-04-13T00:00:00Z", null]]}}


SPACENET_ROADS_NETWORK_DETECTION:
description: |
The SpaceNet 3 dataset was created for the Road Detection and Routing Challenge. It contains more than 8,000 km of road centerlines with detailed attributes, including road type, surface type, and number of lanes. All annotations were digitized from 30 cm GSD WorldView-3 imagery across four cities: Las Vegas, Paris, Shanghai, and Khartoum. This dataset enables the development of algorithms not only for road extraction but also for generating usable routing networks from satellite imagery.
instruments: []
constellation: SpaceNet
platform: SN3
processing:level:
keywords: ["Vegas", "Paris", "Shanghai", "Khartoum", "Road", "Road Network Detection", "SpaceNet", "Chipped Training Dataset"]
eodag:sensor_type:
Comment thread
cauriol marked this conversation as resolved.
Outdated
license: CC-BY-4.0
title: "SpaceNet 3: Road Network Detection"
extent: {"spatial": {"bbox": [[ -115.3075176, 15.5138111999, 121.7498742, 49.0582359], [-115.3075176, 36.1212776997, -115.1513226, 36.2616777], [2.18641139997, 48.9757509, 2.30048639992, 49.0582359], [121.5901692, 31.1978709001, 121.7498742, 31.4277759], [32.4858384, 15.5138111999, 32.5665684, 15.7402062]]}, "temporal": {"interval": [["2015-04-13T00:00:00Z", null]]}}

SPACENET_OFF_NADIR_BUILDING:
description: |
SpaceNet 4 introduced imagery collected at multiple viewing angles (off-nadir), making building footprint extraction significantly more challenging due to distortions, shadows, and perspective changes. The dataset focused on Atlanta, USA, with labeled building footprints provided across different look angles.
instruments: []
constellation: SpaceNet
platform: SN4
processing:level:
keywords: ["Atlanta", "Buildings", "Chipped Training Dataset", "Off-Nadir Buildings", "Off-Nadir", "v1", "SpaceNet"]
eodag:sensor_type:
Comment thread
cauriol marked this conversation as resolved.
Outdated
license: CC-BY-4.0
title: "SpaceNet 4: Off-Nadir Buildings"
extent: {"spatial": {"bbox": [[ 732701.0, 3720189.0, 748901.0, 3744039.0]]}, "temporal": {"interval": [["2016-02-29T00:00:00Z", null]]}}

SPACENET_ROADS_NETWORK_ROUTE_TRAVEL:
description: |
SpaceNet 5 focuses on automated road network extraction and route travel time estimation from satellite imagery. The publicly available dataset includes high-resolution imagery and road labels for two cities: Moscow, Russia, and Mumbai, India. These annotations enable development of algorithms for extracting road networks and generating connected graphs suitable for routing applications.
instruments: []
constellation: SpaceNet
platform: SN5
processing:level:
keywords: ["Moscow", "Road", "Road Network Extraction", "Road Network", "Route Travel Time Estimation", "Route Travel", "SpaceNet"]
eodag:sensor_type:
Comment thread
cauriol marked this conversation as resolved.
Outdated
license: CC-BY-4.0
title: "SpaceNet 5: Automated Road Network Extraction and Route Travel Time Estimation from Satellite Imagery"
extent: {"spatial": {"bbox": [[-43.7753747662, -23.0106794336, 72.89403328656945, 55.746917105720065], [37.62123533463809, 55.60938704097397, 37.78303540799408, 55.746917105720065], [72.7844259905958, 18.885500500608433, 72.89403328656945, 19.0645257518164]]}, "temporal": {"interval": [["2016-02-29T00:00:00Z", null]]}}

SPACENET_ALL_WEATHER_MAPPING:
description: |
SpaceNet 6 (MSAW) introduced multi-sensor data for the first time, combining Synthetic Aperture Radar (SAR) imagery from Capella Space with optical WorldView-2 imagery. The dataset, centered on Atlanta, provides over 48,000 building footprint labels. Training data includes both SAR and optical images, while testing is SAR-only, encouraging robust building detection under all-weather conditions.
instruments: ["Capella-SAR", "WorldView-2"]
constellation: SpaceNet
platform: SN6
processing:level:
keywords: ["Rotterdam", "Buildings", "Building Detection", "SpaceNet", "Multi-Sensor", "All-Weather Mapping", "SAR", "Optical"]
eodag:sensor_type:
Comment thread
cauriol marked this conversation as resolved.
Outdated
license: CC-BY-4.0
title: "SpaceNet 6: All-Weather Mapping"
extent: {"spatial": {"bbox": [[590111.9563080214, 5745470.427302527, 596702.0547804891, 5753195.130788647]]}, "temporal": {"interval": [["2019-08-04T00:00:00Z", "2019-08-23T00:00:00Z"]]}}

# MARK: GENERIC ------------------------------------------------------------------------
GENERIC_COLLECTION:
description:
Expand Down
15 changes: 15 additions & 0 deletions eodag/resources/providers.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5402,6 +5402,21 @@
_collection: EO.ISIMIP.DAT.SOCIO-ECONOMIC-FORCING_ISIMIP3b
GENERIC_COLLECTION:
_collection: '{collection}'
# Spacenet
SPACENET_BUILDINGS_DETECTION_V1:
_collection: EO.SPACENET.DAT.BUILDINGS_DETECTION_V1
SPACENET_BUILDINGS_DETECTION_V2:
_collection: EO.SPACENET.DAT.BUILDINGS_DETECTION_V2
SPACENET_ROADS_NETWORK_DETECTION:
_collection: EO.SPACENET.DAT.ROADS_NETWORK_DETECTION
SPACENET_OFF_NADIR_BUILDING:
_collection: EO.SPACENET.DAT.OFF_NADIR_BUILDING
SPACENET_ROADS_NETWORK_ROUTE_TRAVEL:
_collection: EO.SPACENET.DAT.ROADS_NETWORK_ROUTE_TRAVEL
SPACENET_ALL_WEATHER_MAPPING:
_collection: EO.SPACENET.DAT.ALL_WEATHER_MAPPING


---
!provider # MARK: eumetsat_ds
name: eumetsat_ds
Expand Down
6 changes: 6 additions & 0 deletions tests/units/test_core.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -615,6 +615,12 @@ class TestCore(TestCoreBase):
"CMIP6_CLIMATE_PROJECTIONS": ["cop_cds"],
"TIGGE_CF_SFC": ["ecmwf"],
"UERRA_EUROPE_SL": ["cop_cds", "dedl", "wekeo_ecmwf"],
"SPACENET_BUILDINGS_DETECTION_V1": ["dedl"],
"SPACENET_ALL_WEATHER_MAPPING": ["dedl"],
"SPACENET_BUILDINGS_DETECTION_V2": ["dedl"],
"SPACENET_ROADS_NETWORK_DETECTION": ["dedl"],
"SPACENET_OFF_NADIR_BUILDING": ["dedl"],
"SPACENET_ROADS_NETWORK_ROUTE_TRAVEL": ["dedl"],
GENERIC_COLLECTION: [
"peps",
"usgs",
Expand Down
Loading