University of Tokyo and Kubota Collaborate on Innovative Drone-Based Potato Yield Prediction Technique
Researchers at the University of Tokyo have developed a method to predict potato yield before harvest using drone imagery, machine learning, and a growth curve model. This innovative approach aims to estimate underground tuber biomass, providing valuable insights for agricultural practices.
Methodology of Potato Yield Prediction
The research, conducted by the University of Tokyo Graduate School of Agricultural and Life Sciences in collaboration with Kubota Corporation, utilizes drone-based remote sensing technology. Fields were periodically photographed using drones equipped with RGB and multispectral cameras. The team analyzed various image features, including:
- Plant cover ratio
- Canopy height
- Color indices
- Vegetation indices
A machine-learning model was trained to establish the relationship between these features and the measured underground biomass obtained through sampling. For unharvested plots, the researchers estimated tuber biomass by inputting image features into the model and applying a Gompertz growth curve to predict yield at harvest.
Field Trial Results
The study involved field trials conducted in 2023 and 2024 at the University of Tokyo Field Science Center in Nishi-Tokyo City. The trials included multiple treatment plots with varying planting densities and seed tuber conditions. The researchers achieved a correlation coefficient of 0.8 or higher for tuber biomass estimation and 0.7 or higher for yield prediction using the growth curve. These results indicate that yield can be effectively predicted from pre-harvest data collected via drone technology.
Implications for Smart Agriculture
Potatoes are a significant food crop globally, and traditional yield assessment methods often rely on destructive sampling. The new non-destructive method developed by the research team allows for the assessment of spatial variation across fields. This growth-curve approach is expected to enhance pre-harvest yield forecasting and optimize cultivation management, including determining the optimal timing for harvest. The research was conducted as part of the joint Kubota Todai Lab project.
Further information regarding this research can be obtained from the University of Tokyo.