geoalert / aeronetlib

A Remote Sensing data handling library for Deep Learning

Aeronet

Python library to work with geospatial data

List of content

• Aim and scope
• Modules
• Quickstart example
• Requirements and installation
• Documentation and wiki
• Citing
• License

Aim and scope

As a part of Aeronetlib, which is designed to make it easier for the deep learning researchers to handle the remote sensing data, Aeronet_raster provides an interface to handle geotiff raster images.

Modules and classes

• .raster

  • Band | BandCollection
  • BandSample | BandSampleCollection

• .collectionprocessor

  • CollectionProcessor
  • SampleWindowWriter
  • SampleCollectionWindowWriter

• .visualization

  • add_mask

Quickstart example

Requirements and installation

1. python 3
2. rasterio >= 1.0.0
3. shapely >= 1.7.1
4. rtree>=0.8.3
5. opencv-python>=4.0.0
6. tqdm >=4.36.1

Pypi package: .. code:: bash

$ pip install aeronet[all]

for partial install:

Raster-only .. code:: bash

$ pip install aeronet[raster]

Vector-only .. code:: bash

$ pip install aeronet[vector]

Source code: .. code:: bash

$ pip install git+https://github.com/aeronetlab/aeronetlib

Contributing We accept pull-requests and bug reports at github page

You can use `make build` to build the libraries and `make upload` to update them at pypi (authorization required).

Testing 1. Create and activate virtual environment 2. `make prepare` to install all requirements in the venv 3. `make test` to run all tests

Documentation and wiki

The project wiki contains some insights about the background of the remote sensing data storage and processing and useful links to the external resources. Latest documentation is available at Read the docs

Citing

@misc{Yakubovskiy:2019,
  Author = {Pavel Yakubovskiy, Alexey Trekin},
  Title = {Aeronetlib},
  Year = {2019},
  Publisher = {GitHub},
  Journal = {GitHub repository},
  Howpublished = {\url{https://github.com/aeronetlab/aeronetlib}}
}

License

Project is distributed under MIT License.