A python class for parallelizing functions across the cores of a single computer. Created with large raster processing jobs in mind. Uses multiprocessing module. Made available under a CC BY-NC-SA 4.0 license.
[This project was completed in April 2022] GeocoderPL is an application written in Python, which can be used for geocoding address points in Poland along with the possibility to display basic information about a given address point and the building assigned to this address. GeocoderPL has o form of search engine with three map layers: OpenStreetMap, Google Maps and Stamen's toner map.
In this repository I am sharing a pipeline I have created to ingest, preprocess and finaly create an analysis ready products based on Sentinel 1 and Sentinel 2 images.
This is a script that reads in Landsat-8 data, Esri Sentinel-2 10m land cover time series data and train a random forest classification algorithm to estimate fractional built cover at 30m scale. The trained model can be used to produce fractional land cover for other regions.
This repository presents the Colour Pattern Regression (CPR) algorithm QGIS3 plugin. The code determines the relationship between aerial images and raster maps according to the decomposition into RGB spaces of the aerial images and the calculation of a linear regression with the raster map using three coefficients - one for each RGB space.