Journal of Agronomy and Field Innovations

Background: Triticum aestivum L. (common wheat) is a main grain that remains significant for the overall food security across the world. Conventional ways of estimating yield involves extensive labor, destructiveness, and limitations in spatial scope. Drones equipped with multiple sensors present a non-destructive and quick alternative of monitoring crops.
Objective: The study aimed at evaluating the suitability of NDVI, derived from drones, as a representative of the growing factors of wheat as well as grain harvest.
Method: The study used randomized and complete block design of field experiment by means of multispectral equipped quadcopter flying at significant growth stages. NDVI value was obtained from orthomosaics and getting correlated with plant height, leaf area index, amount of chlorophyll, biomass, and final grain yield through regression and correlation analysis.
Results: NDVI had high temporal variability, being highest at the booting and heading stages, and was significantly associated with yield (R² = 0.86, RMSE = 0.34 t ha⁻¹). NDVI at heading stage was the most effective single-stage predictor of yield.
Conclusion: UAV-derived NDVI is an effective, scalable, and non-destructive method for wheat yield prediction which enables site-specific nutrient management and quick identification of low productivity areas in precision farming systems.