Evaluation Of Sowing Effect And Spatial Variation In Winter Wheat Based On Unmanned Aerial Vehicle Remote Sensing Images

Winter wheat is one of the most important food crops in China.Increasing food demand and limited arable land force agricultural production to develop in a fine and efficient way in order to get the highest economic benefits with minimal resource input.The most important part of agricultural production is sowing.The sowing effect directly affects the efficiency of sowing,the utilization of cultivated land,the management of the field and the grain yield.In addition,the grain yield can be reflected from the side of the crop,and it is of great significance for monitoring the individual characteristics of the crops and the characteristics of the population.The study on the temporal and spatial characteristics of crop growth and the influence of the variable fertilization on the change can provide reference for the cultivation and management of field crops.Taking winter wheat as the research object,using hyperspectral data and UAV remote sensing image as the foundation,combined with the field experimental,the sowing effect,winter wheat growth status,spatial variability and remote sensing and GIS technology combined with sampling survey and agriculture,research and evaluation in the area of winter wheat Fine remote sensing dynamic monitoring index.The main research content and research results of this paper are as follows:(1)Extracting the information of the line center line and the location of the line center point of the winter wheat image based on the early high resolution unmanned aerial vehicle image data.Calculation of the straightness of wheat seeding lines by calculating the deviation from the center line of the center of wheat at the center of the line.The straightness of wheat sowing lines is evaluated by comparing the angle of each line of wheat lines from the baseline evaluated by comparing the angle of each line of wheat lines from the baseline center line.The maximum and minimum value,standard deviation calculation of NDVI in the same study area could be statistical values of each district randomly deployed,in order to evaluate the effect of seeding uniformity.As a result:the mean angle of the non precision sowing area is slightly larger than the mean value of the fine sowing area,and the standard deviation is also slightly higher;It is worth non precision seeding area to deviate from set navigation line 0.0134° by interline calculation;The difference between the standard deviation of the fine sowing and the non precision sowing areas is small,but the mean value of the NDVI is different.(2)The paper is based on the measured canopy hyperspectral data in the early stage of winter wheat growth,we reduced the soil background and other factors as the principle as much as possible so as that we could screen 7 sensitive vegetation index according to the correlation matrix by the software of MATLAB.then optimizing the 40 vegetation indices of double band and multi band based on previous studies,and analysing the correlation between the LAI and the 47 vegetation indices of the sparse winter wheat,at last we established a single variable empirical model and partial least squares regression model to explore the feasibility of hyperspectral data inversion LAI for sparse winter wheat by 10-fold cross-validation for repeating 5 times.As a result.it is feasible to inverse the LAI of winter wheat with hyperspectral data;The effect of water vapor and nitrogen dioxide should be fully considered when using the RVI(1868,1946)-LAI model;PVR(650,550)is the best vegetation index on inversion;Multiple regression model helps to improve the accuracy of inversion.(3)Based on four key growth stages of winter wheat UAV image,NDVI was calculated to evaluate the spatial variability of winter wheat by its variation coefficient and LAI variance variance image.As a result,both NDVI and LAI data showed medium spatial heterogeneity.