link to homepage

Institute of Bio- and Geosciences

Navigation and service

Shape & Color

New affordable methodic approaches and algorithmic solutions for the quantification of quality parameters in vegetables and herbs to speed up horticultural breeding processes

01.12.2019 - 30.11.2022

Projektakronym: Shape & Color
Projekttitel (engl.): Intelligent imaging methods to assess structure and color traits in horticultural plant breeding
Time: 01.12.2019 - 30.11.2022

Förderkennzeichen: 20943 N (IGF-Vorhaben-Nummer)
Topics: horticultural plant breeding, plant phenotyping, image processing
Gefördert von: Bundesministerium für Wirtschaft und Energie (BMWi) / Federal Ministry for Economic Affairs and Energy

In Germany several vegetables crops and herbs are usually grown as field cultivations. Because of the increasing demand for rapid and objective assessment of yield and quality parameters, non-invasive methods need to be evaluated and deployed for the routine characterization of plant productivity traits. Vegetables and herbs, which are featured by well-defined breeding targets and a multitude of parameters relevant for breeding processes, offer new opportunities for non-invasive methods with the potential to considerably reduce the breeding effort, in particular the manual scorings of diverse plant traits.

This joint research project is a cooperation of FZJ-IBG-2, three plant breeding companies, JB Hyperspectral Devices, and the GFPi (Gemeinschaft zur Förderung von Pflanzeninnovation e.V.). The project will focus on the application of non- and minimally invasive (imaging) methods for the quantitative assessment of plant parameters during the selection of horticultural and herbal plant genotypes. Newly developed methods will be put into practice at the project partners sites, namely Rijk Zwaan Marne GmbH, Hild Samen GmbH, and van Waveren Saaten GmbH. Most traits of interest can be described as yield quality parameters that are related to both shape and color appearance. Monitoring such traits over time will contribute to computing growth rates and optimal harvest times. Integration of customary sensor technologies in suitable sensor platforms with a focus on affordable and robust solutions will consider the individual growing scenarios at each breeding company. For the quantitative analysis of measurement and imaging data new algorithmic solutions that follow the breeders’ scoring categories will be developed. The project aims at testing methodologies that give faster and more precise result than conventional manual scorings, thereby reducing the screening effort drastically. In addition, there is potential for new digital proxy parameters for yield quality, so far not yet considered, to be identified as an important outcome of the project.