Context I am replicating the project shown in the City Class project by Roman Kuchokov. The objective is to classify a satellite picture of a small part of a city into one of its zones (such as residential, industrial, etc). I obtained the information about the zoning in Austin to work on the same problem but with a different dataset. Content There are 3666 small satellite images of the whole city of Austin and a csv file which relates the name of each of these images to the corresponding zoning tag. Acknowledgements This project was inspired by Roman Kuchukov's City: https://towardsdatascience.com/cityclass-project-eng-15bc5fcd8e1 The zoning information was obtained in the official city of Austin data portal: https://data.austintexas.gov/Locations-and-Maps/Zoning/5rzy-nm5e Inspiration I want to be able to train a classification model which can take a satellite picture of any city and create a grid where each cell is classified as a type of zone. This may start with zoning classification but can also be tweaked for other urban geo-tagging projects.