Dear Colleagues,

Since the beginning of space Remote Sensing, the basic application of satellite images is for mapping the surface of the Earth that is visible on them. Initially, in the 1970s, the images were visually interpreted. The development of digital techniques has made it possible to try and support this work with algorithms of varying degrees of automation. Initially, basic pixel-based algorithms of supervised and unsupervised classification were applied. The next stage of development was the introduction of object-oriented classification techniques. At the same time, the classification process was increasingly supported by information selected from the existing databases. Currently, machine learning techniques are commonly used, including very advanced deep learning algorithms. More and more often in satellite remote sensing, algorithms are entered that were originally developed for non-image data analysis purposes.

The first global land cover studies were made on the basis of low-resolution images, which resulted from the availability of data and the possibility of their effective processing. Currently, the Earth's surface is recorded with increasing spatial, spectral, and temporal resolution, and simultaneously, the technical capabilities allow for the processing of huge image files and data sets. Mapping large areas such as entire countries, continents, and even the whole world requires the use of dedicated classification approaches, which guarantee a very high degree of automation at each stage of the processing and analysis of satellite data.

This Special Issue invites manuscripts presenting classification approaches and services that are elaborated for global or regional land cover mapping, based on optical or SAR satellite data.

Dr. Stanisław Lewiński
Dr. Garik Gutman
Guest Editors

More Info: https://www.mdpi.com/journal/remotesensing/special_issues/RGLCM