Publication Date:
2020-02-12
Description:
Remote sensing data analysis retrieves spatial-temporal information about the Earth‘s surface from remotely sensed optical and radar images. For this purpose accurate and efficient classification or parameter quantification techniques must be used. Consequently, there exists a long tradition in remote sensing to employ methods and techniques from the field of machine learning. They can be regarded as „universal function approximators“ that are able to link any data in order to derive connections, conclusions and predictions efficiently using different learning strategies. In the following, current research topics of the Remote Sensing section of the GFZ are presented, in which different forms of machine learning are used.
Language:
German
Type:
info:eu-repo/semantics/article
Format:
application/pdf
Format:
application/pdf
