Research and realization of winter wheat yield estimation model based on NDVI index

Abstract

The timely and accurate prediction of crop yield is of great significance to the formulation of national grain policy, the macro control of prices and the development of rural economy. In this paper, the NDVI values were calculated by using the measured spectral reflectance data of Winter Wheat during the whole growth period in 2014, combining with actual measured output, constructing a function model of NDVI index and measured output. The study concluded that the coefficient of determination (R2) of the NDVI index and the measured yield model in the whole growth period was 0.78, the root mean square error is 40.795 (kg/mu), Standard root mean square error is 10.79%. The value of root mean square error of verification model is 49.297 (kg/mu), the value of standard root mean square error is 13.04%. Therefore, the estimation model obtained in this experiment has good reliability, It is feasible that the portable instrument for measuring the parameters of crop growth potential by using the estimation model.