Viertel, PhilippKönig, MatthiasMeyer-Aurich, AndreasGandorfer, MarkusHoffmann, ChristaWeltzien, CorneliaBellingrath-Kimura, SonokoFloto, Helga2021-03-022021-03-022021978-3-88579-703-6https://dl.gi.de/handle/20.500.12116/35697In this work, we will show a use case for visual pollen classification from honey samples. We discuss the current state of the art in pollen analysis, highlight the importance of data quantity and quality, and elaborate on how to transfer promising Deep Learning methods to the analysis of honey samples. A first experiment with a public data set is shown as well as samples from our work-in-progress data set. Our recommendations and methods show which steps are necessary in order to successfully deploy an automated pollen analysis solution for honey products.enDeep LearningMachine LearningPalynologyPollen analysisAutomationText/Conference Paper1617-5468