ASHS 2024 Conference

Potato production is crucial for global food security, and with an increasing demand for food and a diminishing supply of fertile land, there is a need to boost production. Remote sensing technologies, such as high-resolution hyperspectral sensors, have the potential to provide valuable insights into potato growth, yield, and quality. Narrow spectral bands captured by these sensors are directly linked to biophysical parameters and can accurately estimate crop parameters. Recent studies have utilized hyper-spectral imagery acquired from proximal sensor such as ASD FieldSpec to estimate various crop parameters and yield. The results from these studies are promising, indicating that hyper-spectral sensors have the potential to improve crop management and agricultural practices. Moreover, the integration of remote sensing data with advanced analytical techniques, such as machine learning, helps in accurately estimating yield and yield components. In this study we are testing two machine learning such as PLSR and RF to predict biomass and N uptake in-season. Results from this indicate that PLSR performs better in predicting biomass and N uptake in potatoes. Moreover, yield can be best estimated at tuber bulking stage.