Python implementation of some simple algorithms for pansharpening remote sensing images

This repository contains a function which takes as input multispectral and panchromatic remote sensing images, and outputs a pansharpened version of the multispectral image using either one of the following algorithms:

• Simple Brovey
• Simple Mean
• ESRI
• Brovey

This function is mainly adapted from the source code written by Vladimir Osin in his Kaggle Kernel 'Panchromatic Sharpening' in the link https://www.kaggle.com/resolut/panchromatic-sharpening.

The implementation in this repository is mainly to adapt his code to accept raster multispectral and panchromatic files as inputs instead of the images in their array form, generalize the code for accepting multispectral images with differing number of bands and band order, and also to generate the raster pansharpened image file as well.

Requirements:

• cv2
• gc
• numpy
• rasterio

A zoomed - in portion of a GeoEye - 1 sample image (courtesy of European Space Agency) is used to illustrate the Brovey pansharpening technique.

Test Multispectral Image (Courtesy of European Space Agency):

Test Panchromatic Image (Courtesy of European Space Agency):

Pansharpened Image (Simple Brovey):

Pansharpened Image (Brovey with weight of 0.15):