Remote Sensing Based Land Cover Classification Using Machine Learning and Deep Learning: A Comprehensive Survey

Since the 1990s, remote sensing images have been used for land cover classification combined with Machine Learning algorithms. The satellites and airborne sensors pass over a specific point of land surface periodically, it is possible to assess the change in land cover over a long time. With the advent of Machine Learning (ML) methods, automated land cover classification has been at the center of research for the last few decades. From 2015 forward, a technical shift has been noticed with the emergence of several branches of Neural Networks (NN) and Deep Learning (DL). This paper examines current practices, problems, and trends in satellite image processing. This formal review focused on the summarization of major classification approaches from 1995. Two dominant research trends have been noticed in automated land cover classification, e.g., per pixel and subpixel analysis. Classical machine learning algorithms and deep learning methods are mainly used for per-pixel analysis, whereas fuzzy logic algorithms are used for sub-pixel analysis. The current article includes the research gap in automated land cover classification to provide comprehensive guidance for subsequent research direction.