Related Experiment Video
Updated: Dec 8, 2025
06:28
Author Spotlight: Unraveling Plant Responses to Abiotic Stresses Using the PlantScreen Robotic Platform
Published on: June 7, 2024
2.4K
Prediction of End-Of-Season Tuber Yield and Tuber Set in Potatoes Using In-Season UAV-Based Hyperspectral Imagery and
Chen Sun1,2, Luwei Feng1, Zhou Zhang1
1Biological Systems Engineering, University of Wisconsin-Madison, Madison, WI 53706, USA.
Sensors (Basel, Switzerland)
|September 19, 2020
Summary
Unmanned aerial vehicle hyperspectral imaging and machine learning accurately predict potato tuber yield and tuber set. This technology aids growers in optimizing irrigation for sustainable agriculture.
Area of Science:
- Agricultural Science
- Remote Sensing
- Machine Learning
Background:
- Potato is a major global food crop, necessitating efficient yield prediction for sustainable management.
- Precision agriculture utilizes Unmanned Aerial Vehicles (UAVs) for enhanced crop monitoring.
- Hyperspectral imaging offers superior spectral detail compared to natural color and multispectral data for crop trait modeling.
Discussion:
- This study evaluated six machine learning models for predicting potato tuber yield and tuber set using UAV-based hyperspectral data.
- Models were tested on research plots with varying irrigation levels.
- Multi-temporal hyperspectral data improved prediction accuracy for both tuber yield and tuber set.
Key Insights:
- Tuber set prediction was more accurate than tuber yield prediction.
- Ridge regression yielded the best results for tuber yield prediction (R² = 0.63).
- Ridge regression and Partial Least Square Regression (PLSR) showed comparable performance for tuber set prediction (R² = 0.69).
Outlook:
- Hyperspectral imagery combined with machine learning shows significant potential for potato growers.
- This approach can assist in optimizing irrigation management strategies.
- Further research can refine these models for broader agricultural applications.
