AI-based satellite image classification: an analysis of different algorithms on PlanetScope and sentinel imagery

Abstract

This research aims to explore the potential of machine-learning algorithms in land classification and effectively analyse and visualize the results through the development of an application. The study area and different satellite systems used for data acquisition, and various machine-learning algorithms for satellite image classification are explored. The research also delves into Geographic Information Systems (GIS) concepts and methods of classification, including different algorithms used for satellite image processing. The analysis and interpretation of the results are covered in detail, using open-source software to differentiate between Sentinel-2 and Planet da-ta sets visually. The developed application provides a comprehensive and effective visualization of the results. Overall, this study provides a comprehensive understanding of the potential of machine learning algorithms in land classification and their effectiveness in analysing and visualizing results.