The task is to determine NDVI values, reproject, scale, and output a geotiff based on another 8-band geotiff.

For more information, please refer to this explanation on wikipedia.

To run the script, libraries are required.

• GDAL
• rasterio
• numpy

These can be installed in a new virtual env or found in a pre-existing docker image Development Seed offers.

The script can be ran from the command line or included in a project as an import.

The python script can be ran from the command line. Two arguments are required, an input and output file path, name.

python detect.py input.tif output.tif

Optional arguments include --crs and --range.

Including --crs tells the app to reproject the output to the given projection.

Including --range tells the app to clamp the output NDVI values to the given range.

From README directory

python ndvidetect/detect.py sample/sample.tif output/output.tif --crs EPSG:4326 --range_min 0.2 --range_max 0.5

The geotif included (output/output.tif) was generated by the above command.

Once the repo is built/installed with python's setup tools (python install .), you can access key functions to calculate NDVI values of a geotiff.

from ndvidetect import evaluate_band_vegetation, evaluate_geotiff_vegetation

Once imported, these two functions can be used to determine NDVI values.

from ndvidetect import evaluate_geotiff_vegetation
    
if __name__ == "__main__":
    fn_in = 'sample.tif'
    fn_out = 'output.tif'
    evaluate_geotiff_vegetation(fn_in, fn_out, crs='EPSG:4326', range=[0.3, 0.6])

Further documentation is included on the functions.