The objective was to develop imaging methods to estimate the root biomass of robustly growing crop plants, in order to determine the amount of carbon and nitrog

The objective was to develop imaging methods to estimate the root biomass of robustly growing crop plants, in order to determine the amount of carbon and nitrogen remaining in the soil after harvest, based on above-ground plant parts. The relationship between root biomass and above-ground traits was studied in more detail for faba bean, caraway, and industrial hemp. The best predictors of root biomass were combinations of stem base thickness and seed yield or plant height. 
A machine learning model was also developed to estimate root biomass using whole plant samples. 
Drone imaging focused on analyzing correlations between indicators such as the Enhanced Green Index (ExG) and plant height. 
Yield Systems developed applications for crop imaging. Geo-Work Oy tested ground-penetrating radar (GPR), which, based on the trial, may be suitable for research use in the future. In field conditions, soil properties currently limit its effectiveness.
Based on the results, a calculator prototype was created to estimate root biomass per hectare and the corresponding carbon and nitrogen amounts. Accuracy can be improved with annual reference datasets.
The project introduced several innovations, including:
•A resource-efficient method for sampling whole plants
•Identification of stem base thickness, seed yield, and plant height as key parameters
•Support for device development aimed at root biomass estimation
•Utilization of drone technology and mobile applications
•Potential use of GPR in container or pot cultivation
Conclusion: Root biomass can be estimated based on above-ground plant characteristics, but improving reliability requires further development and more data.