top of page

Application Of Small UAVs In Canola Yield Prediction Using Yellowness Index

Abstract: Recent developments in technologies for imaging plants using small Unmanned Aerial Vehicles (s-UAV) make it possible to effectively collect various crop phenotyping data and to predict crop yield via radiometric data. Canola flowers indeterminately over a protracted period with bright yellow flowers. Yellowness of canola petals may be a critical part of a canopy-level signal and a good predictor of seed number and therefore seed yield. Our objectives were 1) to evaluate which wavebands of light may be useful/sensitive for estimation of yellowness, and 2) to determine whether the yellowness index of canola canopy during flowering could be a sensitive predictor of seed number and seed yield. There were 56 canola genotypes grown at Saskatoon, SK, Canada in 2016. Hand-held spectroradiometer was used to collect reflectance values from the ground during flowering. Meanwhile, aerial imagery with geometric and radiometric corrections was collected using an s-UAV mounted with a multispectral and RGB digital cameras. We expect that the ratio of blue band over green band will be a useful index for representing canola yellowness, which is related to canola flowing intensity. The yellowness index integrated over time should be a good predictor of seed number and thus canola yield.

Authors:

Ti Zhang, Hema Sudhakar Duddu, Menglu Wang, Steve Shirtliffe

Institutions:

University of Saskatchewan

bottom of page