Estimation of biophysical parameters in wheat crops in Golestan province using ultra-high resolution images

ABSTRACT

Due to various factors which impress wheat growth variability, we need short–term monitoring of biophysical parameters using ultra–high resolution images. These provide an ability to monitor crops at the individual plant level. Two flight missions were carried out at altitude of 40 m with a commercial quad copter and a commercial camera. The images were taken before and after tillage over an 8.8 ha field. The 2 cm orthoimages and surface models were generated using photogrammetric software. Then, the image variables including ratio of the blue and green band, ratio of the red and blue band,ratio of the red and green band, plant height were extracted from orthoimages and surface models. Field measurement included the leaf area index, plant height and biomass for 15 plots each of area 1 m × 1 m. Due to a linear relationship between the biophysical parameters and image variables, it was used a multivariate regression model for modelling. The model using the image variables resulted in coefficient of determination (R²) of 0.95 and the lowest error measures (RMSE = 0.24). The results show that ultra–high resolution images can be used for monitoring of biophysical parameters in wheat crops but it is limited for large area.