Remotior Sensus (which is Latin for "a more remote sense") is a Python package that allows for the processing of remote sensing images and GIS data.
Remotior Sensus is developed by Luca Congedo.
- Website: https://fromgistors.blogspot.com/p/remotior-sensus.html
- Documentation: https://remotior-sensus.readthedocs.io
- Source code: https://github.com/semiautomaticgit/remotior_sensus
- Bug reports: https://github.com/semiautomaticgit/remotior_sensus/issues
The main objective is to simplify the processing of remote sensing data through practical and integrated APIs that span from the download and preprocessing of satellite images to the postprocessing of classifications and GIS data. Basic dependencies are NumPy, SciPy for calculations, and GDAL for managing spatial data. Optionally, Matplotlib is used to display spectral signature plots.
The main features are:
- Search and Download of remote sensing data such as Landsat and Sentinel-2.
- Preprocessing of several products such as Landsat and Sentinel-2 images.
- Processing and postprocessing tools to perform image classification through machine learning, manage GIS data and perform spatial analyses.
- Parallel processing available for most processing tools.
WARNING: Remotior Sensus is still in early development;
new tools are going to be added, tools and APIs may change,
and one may encounter issues and bugs using Remotior Sensus.
Most tools accept raster bands as input, defined through the file path.
In addition, raster bands can be managed through a catalog of BandSets, where each BandSet is an object that includes information about single bands (from the file path to the spatial and spectral characteristics). Bands in a BandSet can be referenced by the properties thereof, such as order number or center wavelength.
Multimple BandSets can be defined and identified by their reference number. Therefore, BandSets can be used as input for operations on multiple bands such as Principal Components Analysis, classification, mosaic, or band calculation.
In band calculations, alias name of bands based on center wavelength (e.g. blue, red) can be used to simplify the structure of calculation expression.
Most tools are designed to run in parallel processes, through a simple and effective parallelization approach based on dividing the raster input in sections that are distributed to available threads, maximizing the use of available RAM. This allows even complex algorithms to run in parallel. Optionally, the output file can be a virtual raster collecting the output rasters (corresponding to the sections) written independently by parallel processes; this avoids the time required to produce a unique raster output. Most tools allow for on the fly reprojection of input data.
Remotior Sensus optional dependencies are PyTorch and scikit-learn, which are integrated in the classification tool. to allow for land cover classification through machine learning. The aim is to simplify the training process and development of the model.
Remotior Sensus requires GDAL, NumPy and SciPy for most functionalities. Also, scikit-learn and PyTorch are optional but required for machine learning. Optionally, Matplotlib is used to display spectral signature plots.
It is recommended to install Remotior Sensus using a Conda environment.
$ conda install -c conda-forge remotior-sensus scikit-learn pytorchRemotior Sensus is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version. Remotior Sensus is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details. You should have received a copy of the GNU General Public License along with Remotior Sensus. If not, see https://www.gnu.org/licenses/.
For more information and tutorials visit the official site
From GIS to Remote Sensing