In the AI project for wood species identification in fiber materials (AI-Wood_ID), one of the most important factors is to generate as much training data as possible for machine learning. Specifically, these are microscopic images of whole slides with stained fibrous materials of different wood species. With their help, the AI is trained to recognize the cells they contain.
A new light microscope has recently been installed in the fiber laboratory, into which 100 slides can be inserted at once. These are then scanned automatically over the entire surface of the slides in several focus layers. In 24/7 operation, the image data volumes required for the project can thus be collected extremely efficiently and forwarded to the project partners of the Fraunhofer Institute ITWM.
To the Thünen project page: Automated image recognition for wood species identification