rasdaman newsletter 04/2020


Datacubes meet AI: A Game Change for Crop Mapping and Land Cover

A new era of land cover and crop mapping has begun with Coramaps GmbH, a spinoff company located in Darmstadt. Coramaps allows access to near real-time information for very large areas on land cover by using satellite radar data. This fully computer-based analysis and interpretation works on the basis of an AI-based algorithm in combination with rasdaman as a database.

New opportunities to capture our planets’ surface and atmospheric condition arise with the new era of satellite-based earth observation. Especially synthetic aperture radar (SAR) sensors deliver reliable, dense data time series. Coramaps is a Startup emerged from Technische Universität Darmstadt and founded in December 2019. It has a special focus on an AI-based interpretation of satellite based radar data. In this context Coramaps extracts detailed and rapid information about agricultural areas with regard to the follwing questions: What grows where, how much of it and in what quality?

“The centerpiece of our processing environment is the rasdaman database, which allows for fast and flexible management and accessibility of the data. For large multi-temporal data stacks, this database enormously increases the data mining potentials. rasdaman became our tool of choice after a sorrow evaluation of different options where no one was as fast and efficient as rasdaman. It performs in perfection, embedded in our image analysis procedures which are written mainly in Python and R”, Dr. Damian Bargiel says.

He and Dr. Markus Huhn constitute the management board. “rasdaman is fast, flexible and efficient, no matter which kind of satellite data! The supported OGC standards, especially WCPS, increases rasdaman's interoperability and allowed a flawless integration of rasdaman into our image processing procedures while providing stellar performance.”

Read more