This repo provides tools for using napari to work with remote sensing data files:
- image_labeller: Launch the napari image viewer/labelling tool on an Sentinel 3 / SLSTR scene or slsitr-lite scene
- netcdf_viewer: Launch napari to view 2D gridded data from any netcdf4 file
For more information on napari see: https://napari.org/
- conda or miniconda (recommended)
- git
Install the napari python package, preferably into a new conda environment (see https://napari.org/stable/getting_started/installation.html#install-as-python-package for instructions for your platform)
On linux (Ubuntu 24.04), the following installation process has been tested:
conda create -n napari-env python=3.12
conda activate napari-env
pip install napari[all]
Check the version of napari that has been installed (assuming your conda environment is called napari-env):
$ conda activate napari-env
(napari-env) $ napari --version
napari version 0.7.0
Check that the napari tool UI can open:
(napari-env) $ napari
Create a new folder and clone this repo into it
mkdir napari_slstr_install
cd napari_slstr_install
git clone https://github.com/surftemp/napari-slstr.git
Install napari-slstr image_labeller and netcdf_viewer commands from the repo:
cd napari-slstr
pip install -e .
Check that the image_labeller tool will open, requires an unzipped Sentinel3 SLSTR scene, eg
image_labeller ~/Projects/napari/S3A_SL_1_RBT____20170122T094902_20170122T095202_20181004T064548_0179_013_264______LR1_R_NT_003.SEN3 ./resources/example.cfg
To install updates to the image_labeller and netcdf_viewer tools:
cd napari_slstr_install
git pull
In the resources/example.cfg configuration file, all channels are shown, with nadir and oblique views, plus a false colour view constructed from bands S3 (red), S2(green) and S1(blue).
image_labeller <path-to-SLSTR-scene-folder> ./resources/example.cfg
When viewing an slstr-lite scene, use configuration file (or one based on) ./resources/example_slstr_lite.cfg
updated label files are written into the scene folder when the user closes Napari (select Exit from the File menu).
It is recommended to periodically back up label files.
The netcdf_viewer tool uses command line options to control which files and variables are to be displayed, and how to display them. For example:
netcdf_viewer.py [-h] [--x-dim X_DIM] [--y-dim Y_DIM] path layer [layer...]
netcdf_viewer LC08_L1TP_205025_20230420_20230429_02_T1.nc rgb(B4:B3:B2) B11:275:295:coolwarm --y-dim=nj
- the netcdf file is
LC08_L1TP_205025_20230420_20230429_02_T1.nc - two layers are requested:
- the first layer is a rgb false colour composed from r=B4,g=B3,b=B2
- the second layer displays B11 using the coolwarm colour map between 275K and 295K
- use dimension nj as the y axis
Specify multiple paths and their layers using:
netcdf_viewer.py [-h] [--x-dim X_DIM] [--y-dim Y_DIM] path layer [layer...] path layer [layer...]
--x-dim and --y-dim specify the data dimensions to display on x and y axes. Prefix the dimension name with a minus sign to flip the axes.
Layers are specified using the following (hopefully self-explanatory) notation:
Use rgb(r-var-name:g-var-name:b-var-name) to specify 3 bands to compose a false colour layer
or
Use var-name[:min-value:max-value[:colour-map]] to plot a single band, with optional min/max and colour map.
The following colour maps are available:
"GrBu", "GrBu_d", "PiYG", "PuGr", "RdBu", "RdYeBuCy", "autumn", "blues", "cool", "coolwarm", "cubehelix", "diverging", "fire", "gist_earth", "gray", "gray_r", "grays", "greens", "hot", "hsl", "hsv", "husl", "ice", "inferno", "light_blues", "magma", "orange", "plasma", "reds", "single_hue", "spring", "summer", "turbo", "twilight", "twilight_shifted", "viridis", "winter"