CC-BY-NC-4.0

power-tower-training-dataset-for-deep-learning-model

NEO (https://www.neo.nl/home-int/) is a leading company in the Earth Observation (EO) business based in the Netherlands.  Established in 1996, NEO has 25 years track record in providing information services starting from the changes in our habitat. From the changes we detect, we built value-added services supporting both public and private organizations.

NEO's EO-based solutions cover a wide range of applications including but not limited to: Energy and Asset management, Infrastructure monitoring, Environmental monitoring, Agriculture monitoring, as well as Land and water. 

With the strong expertise in geo-information/remote sensing and the patented and certified solution (ISO9001 and ISO27001) SignalEyes® for monitoring with earth observation data, NEO translates geodata into value-added services for the clients.

NEO serves over 250 customers in The Netherlands, within the European Union and internationally. Its services empower our customers to make better decisions, reduce costs, increase safety and prevent errors, to comply with regulations and to manage our environment better.

The dataset contains geojson files of power tower location with attributes about coordinates and associated image tiles in .tif format. The attributes in the json file are as following:
visibility: everything with NULL is visbible, poor and very poor are hard to see in mapbox imagery
set: everything with "test" in this column, was part of our test data
zoom_level: the slippy map zoom_level (18 everywhere)
tile_size: the size of the used tiles (512 everywhere)
lat: the latitude of the tower in EPSG:4326
long: the longitude of the tower in EPSG:4326
tile_name: the slippy map tile name (x_y) is also the filename!
x: the x-pixel coord of the tower in the image
y: the y-pixel coord of the tower in the image
geometry:  the geomerty attribute for this tower in EPSG:4326

power tower DL training dataset - India

Yemen image tiles.zip

The dataset contains geojson files of power tower location with attributes about coordinates and associated image tiles in .tif format. The attributes in the json file are as following: visibility: everything with NULL is visbible, poor and very poor are hard to see in mapbox imagery set: everything with "test" in this column, was part of our test data zoom_level: the slippy map zoom_level (18 everywhere) tile_size: the size of the used tiles (512 everywhere) lat: the latitude of the tower in EPSG:4326 long: the longitude of the tower in EPSG:4326 tile_name: the slippy map tile name (x_y) is also the filename! x: the x-pixel coord of the tower in the image y: the y-pixel coord of the tower in the image geometry: the geomerty attribute for this tower in EPSG:4326