Shapescience – AI for morpholigcally based fruit variety recognition

At a glance
- Project leader : Prof. Dr. Thomas Ott
- Project team : Dr. Martin Rerabek, Anna Wróbel
- Project status : ongoing
- Funding partner : Federal government (Bundesamt für Landwirtschaft BLW)
- Project partner : Eidgenössische Technische Hochschule Zürich ETH, Realisation Schmid
- Contact person : Thomas Ott
Description
Pomological and molecular genetic methods have been the tools of choice to describe fruit varieties and their characteristics. Thanks to the possibility of using 3D scanning technology and machine learning, there is a revival of phenotyping, i.e. the typification of fruits based on their appearance. The Shapescience project attempts to use the potential offered by state-of-the-art machine learning approaches for phenotyping in selected areas of pomology and variety breeding.