Interoperability
Image processing software
Some python image processing software packages
organize arrays differently than rasterio. The interpretation of a
3-dimension array read from rasterio
is:
(bands, rows, columns)
while image processing software like scikit-image
, pillow
and matplotlib
are generally ordered:
(rows, columns, bands)
The number of rows defines the dataset’s height, the columns are the dataset’s width.
Numpy provides a way to efficiently swap the axis order and you can use the following reshape functions to convert between raster and image axis order:
>>> import rasterio
>>> from rasterio.plot import reshape_as_raster, reshape_as_image
>>> raster = rasterio.open("tests/data/RGB.byte.tif").read()
>>> raster.shape
(3, 718, 791)
>>> image = reshape_as_image(raster)
>>> image.shape
(718, 791, 3)
>>> raster2 = reshape_as_raster(image)
>>> raster2.shape
(3, 718, 791)