navisetr.blogg.se - Spacenet tutorial

SPACENET TUTORIAL PATCH

Images are grouped into sets of five, each of which have the same setId.Kaggle - Draper - place images in order of time I believe there was a problem with this dataset, which led to many complaints that the competition was ruined.Rating - medium, most solutions using deep-learning, many kernels, good example kernel.Interview with 1st place winner who used segmentation networks - 40+ models, each tweaked for particular target (e.g.10 Labelled classes include - Buildings, Road, Trees, Crops, Waterway, Vehicles.WorldView 3 - 45 satellite images covering 1km x 1km in both 3 (i.e.Rating - medium, many good examples (see the Discussion as well as kernels), but as this competition was run a couple of years ago many examples use python 2.FastAI Multi-label image classification.1st place winner interview - used 11 custom CNN.12 classes including - cloudy, primary + waterway etc.3-5 meter resolution GeoTIFF images from planet Dove satellite constellation.Kaggle - Amazon from space - classification challenge A list if general image datasets is here. Kaggle hosts several large satellite image datasets ( > 1 GB). Shuttle Radar Topography Mission (digital elevation maps) Imagery on GCP, see the GCP bucket here, with imagery analysed in this notebook on Pangeo.8 bands, 15 to 60 meters, 185km swath, the temporal resolution is 16 days.Long running US program -> see Wikipedia.

This is a multi-label dataset with 43 imbalanced labels.

SPACENET TUTORIAL PATCH

The image patch size on the ground is 1.2 x 1.2 km with variable image size depending on the channel resolution. bigearthnet - The BigEarthNet is a new large-scale Sentinel-2 benchmark archive, consisting of 590,326 Sentinel-2 image patches.

eurosat - EuroSAT dataset is based on Sentinel-2 satellite images covering 13 spectral bands and consisting of 10 classes with 27000 labeled and geo-referenced samples.

The dataset is distributed over 42 cities across different continents and cultural regions of the world.

so2sat on Tensorflow datasets - So2Sat LCZ42 is a dataset consisting of co-registered synthetic aperture radar and multispectral optical image patches acquired by the Sentinel-1 and Sentinel-2 remote sensing satellites, and the corresponding local climate zones (LCZ) label.

Example loading sentinel data in a notebook.

Paid access via sentinel-hub and python-api.

13 bands, Spatial resolution of 10 m, 20 m and 60 m, 290 km swath, the temporal resolution is 5 days.

As part of the EU Copernicus program, multiple Sentinel satellites are capturing imagery -> see wikipedia.

For more Worldview imagery see Kaggle DSTL competition.

Package of utilities to assist working with the SpaceNet dataset.

cloud_optimized_geotif here used in the 3D modelling notebook here.

SpaceNet dataset on AWS -> see this getting started notebook and this notebook on the off-Nadir dataset.

Various datasets listed here and at awesome-satellite-imagery-datasets.

Warning satellite image files can be LARGE, even a small data set may comprise 50 GB of imagery.

Long list of satellite missions with example imagery.

random forests, stochastic gradient descent) are also discussed, as are classical image processing techniques. To a lesser extent Machine learning (ML, e.g. This document primarily lists resources for performing deep learning (DL) on satellite imagery. Robmarkcole/satellite-image-deep-learning Resources for deep learning with satellite & aerial imagery