Date of this Version
Biosystems Engineering 189 (2020), 24-35
Accurate crop yield estimation is important for agronomic and economic decision-making. This study evaluated the performance of imagery data acquired using a unmanned aerial vehicle (UAV)-based imaging system for estimating yield of maize (Zea mays L.) and the effects of variable-rate nitrogen (N) application on crops. Images of a 27-ha maize field were captured using a UAV with a consumer-grade RGB camera flying at ~100 m above ground level at three maize growth stages. The collected sequential images were stitched and the Excess Green (ExG) colour feature was extracted to develop prediction models for maize yield and to examine the effect of the variable-rate N application. Various linear regression models between ExG and maize yield were developed for three sample area sizes (21, 106, and 1058 m2). The model performance was evaluated using coefficient of determination (R2), F-test and the mean absolute percentage error (MAPE) between estimated and actual yield. All linear regression models between ExG and yield were significant (p ≤ 0.05). The MAPE ranged from 6.2 to 15.1% at the three sample sizes, although R2 values were all <0.5. Prediction error was lower at the later growth stages, as the crop approached maturity, and at the largest sample level. The ExG image feature showed potential for evaluating the effect of variable-rate N application on crop growth. Overall, the low-cost UAV imaging system provided useful information for field management.