Institute of Geography
Mapping the long-term landscape changes over large areas by combining the historical maps and the satellite data
Name of the supervisor
Mgr. Róbert Pazúr, PhD.
Prírodovedecká fakulta UK
Sources like orthophotos and topographic maps go back to the beginning of the 20th century. This PhD aims to use current development in image processing and remote sensing to map the historical development of landscapes over large areas. As historical data like topographic maps, aerial photos and Corona, Landsat and Sentinel satellite imagery will be used in this project. Nowadays, processing of historical data is very time consuming and usually requires manual identification of the landscape features over the entire area of mapping. A perspective solution here might be using the current technological advancements of image processing (e.g., Keras, Tensorflow) like the object-based classification and neural network classification. Assuming successful training of the image classification framework may substantially speed-up the classification process and allow to explore the historical development of the landscapes over large areas. A successful candidate should be fluent in English, be capable of working in the field of spatial analysis, remote sensing and data science. This PhD position offers a great way of joining international research projects.