mysnap

An experimental package with simple utilities for quickly importing and handling ESA SNAP toolbox products in BEAM-DIMAP format.

Package available from GitHub, documentation available here.

Quick Start

Install package from the git repository:

$ pip install git+https://github.com/nunorc/mysnap@master

Open a product, build images by stacking bands, and split in tiles.

import mysnap

# product filename in beam-dimap format
filename = '~/Products/subset_2021_resampled.dim'

# load product from file
product = mysnap.Product(filename)

# check rasterized size of image and number of bands
product.meta.ncols, product.meta.nrows, product.meta.nbands  # 2312, 2312, 43

# get list of bands
product.bands  # ['b1', 'b2', 'b3', ...

# access individual band
product.b1.name    # 'b1'
product.b1.desc    # 'B1 (443.0)'
product.b1.array   # array([[1731., 1731., 1731., ...
product.b1.array.shape   # (2312, 2312)

# build a stacked array of selected bands
bands = ['b2', 'b3', 'b4', 'b5', 'b6', 'b7', 'b8', 'b8a', 'b11', 'b12']
arr = product.stack(bands, normalize=mysnap.normalize)
arr.shape   # (2312, 2312, 10)

# split stack in 64x64 tiles
tiles = mysnap.tiles(arr, size=64)
tiles.shape   # (1296, 64, 64, 10)

# build a rgb image
bands = ['b4', 'b3', 'b2']
rgb = product.stack(bands, quantiles=[0.025, 0.975], max=3000, normalize=mysnap.normalize)

Acknowledgments

Raster data imported with rasterio. Thank you to ESA, the authors of the SNAP Toolbox, and upstream packages and products.