Hyperspectral Apple Disease Detection
Centro di Sperimentazione Laimburg, together with BIOMETiC, the Free University of Bozen/Bolzano, and Eurac Research, is collaborating on the HIPPA project, which started at the end of 2023 and is now in its middle phase. The aim is to identify early postharvest physiological disorders and diseases in apples using hyperspectral images acquired by special cameras. Predicting non-manifest disorders will reduce wastage by identifying the best time to sell apples.
Early detection of physiological damage in apples without needing human intervention. This is what the HIPPA project, funded by the European Regional Development Fund (ERDF), in which BIOMETiC is collaborating with the Laimburg Experimentation Center, Free University of Bolzano, and Eurac Research, aims to do. While it is possible to easily eliminate fruits that show external damage, recognizing internal defects is still a challenge. Launched in 2023, the project is investigating a methodology for early detection of pathologies and physiological disorders that may arise in fruit during cold storage, employing state-of-the-art tools and machine learning techniques.
Identifying damaged apples before the appearance of visible symptoms
Through hyperspectral imaging, researchers study the physiological state of fruits, analyzing pathologies and predicting symptoms not yet visible. The hyperspectral imaging technique also allows information to be obtained in the non-visible spectrum, thus increasing the amount of data that can be obtained from a single photograph. The hyperspectral images collected as part of the HIPPA Project will be used to train an artificial intelligence that will be able to understand which fruits might develop pathologies or physiological damage.
Predicting non-manifest disorders will reduce wastage by identifying the best time to sell apples. The ultimate goal is to promote environmentally and economically sustainable apple production and distribution.
Hyperspectral imaging in the service of agriculture
Hyperspectral imaging, already used in other fields such as art history, particularly in the analysis of artworks, is now emerging in the food industry as well. Developed by NASA in the 1970s and 1980s, this technique makes it possible to obtain data on photographed subjects even outside the spectrum of light visible to the naked eye. In the HIPPA Project, a hyperspectral camera operating in the visible and infrared allows for the collection of apple-specific data, from which the research team can derive information about the physiological state of the fruit.
Article author: Debora Lamcja – Laimburg Experimentation Center
Image Credits: © Laimburg Research Centre/Ivo Corrà /Agnese Martinelli
If you would like to find out more about this project, you can check out the following links:
PROJECTS – Vision Computing and Learning Lab (unibz.it)
efre-fesr Südtirol-Alto Adige / Free University of Bozen-Bolzano (unibz.it)
