Estimation of Density and Height of Winter Wheat Varieties Using Unmanned Aerial Vehicles Images

Abstract

In order to quickly and accurately obtain density and height information of winter wheat varieties, it is of great practical significance for the growth monitoring of new wheat varieties. In actual production, the plant density and height are mainly obtained by manual measurement, which is inefficient, time-consuming and laborious. Therefore, the winter wheat were extracted coverage based on unmanned aerial vehicles (UAV) images at seedling stage, the relationship between coverage and plant density were investigated. Moreover, the high-definition digital images of winter wheat varieties at 4 growth stages including jointing, booting, flowering and grain filling stages were obtained. The digital orthophoto model (DOM) and digital surface model (DSM) of winter wheat varieties was generated in combination with the ground control points. The estimation model of plant height at the four growing stages were established. Based on the ground measured plant height (H) of new wheat varieties, the plant height of new wheat varieties extracted by DSM was verified. The results showed that the coverage of new wheat varieties extracted from the UAV images at seedling stage was highly correlated with the measured plant density, and the coefficient of determination (R2) was 0.82. The new wheat varieties H extracted by DSM was significantly correlated with the measured H, and the fitted R2 and root mean square error (RMSE) of the predicted plant height and the measured value were 0.96 and 6.32 cm, respectively. It indicated that the use of UAV images to predict the plant density and plant height of new wheat varieties has good applicability, and can provide technical reference for the monitoring of wheat phenotypic information in the future.