header_bild

DiPredict

AI-based optimization of selection under drought stress in wheat breeding


Term

2024-04-01 bis 2027-12-31

Project management

  • Andreas, Stahl
  • Gwendolin, Wehner


Responsible institute

Institut für Resistenzforschung und Stresstoleranz


Project preparer

  • Andreas, Stahl
  • Sebastian, Warnemünde
  • Gwendolin, Wehner

Cooperation partner

  • Martin-Luther-Universität Halle-Wittenberg


Overall objective of the project

The project aims to optimize sensor-based evaluation of drought stress tolerance in wheat and make it high-throughput. This initially involves AI modeling of water use efficiency using hyperspectral camera sensors in a controlled stress environment (Plantarray, JKI Quedlinburg). This will be followed by the transfer of the findings obtained in this way via high-throughput drone technology to plant breeding field trials at various locations in Saxony-Anhalt. In addition, AI-based methods will be developed to predict the yield structure of different wheat genotypes from drone images taken by various sensors (RGB, hyperspectral, LiDAR). To this end, new and advanced AI systems for sensor data analysis will be developed, specially adapted to the particular requirements of outdoor environments.


Funder

Europäische Union