VALIDATION OF DATA OBTAINED AFTER FIELD SENSING USING UAV FOR MANAGEMENT OF FUTURE CROPS

Abstract

The use of UAVs in crop production is one of the priority areas for increasing yields and maintaining soil fertility. Consideration of changes in the illumination of the objects under study is critical for spectral monitoring when using vision devices. To carry out an atmospheric correction, modern serial spectral monitoring complexes must have a standard anti-aircraft sensor fixed from the top of the UAV. With such a solution, a situation is quite possible when, during the flight, due to the inclination of the aircraft, the sensor systems are at an angle to the horizon, which will change, in particular, when the device is deployed, which can lead to false results. For budget vehicles created for purely monitoring purposes, the use of specialized sensors to control the angle of attack of the aircraft is considered a dubious decision for economic reasons. An alternative to specialized sensors can be software tools. Currently, there are no methods for the programmatic assessment of the suitability of automatic atmospheric correction of spectral data, the development of which was the purpose of the work. Field studies were carried out on October 30, 2019, in the Boryspil region on industrial crops of winter rapeseed and wheat. The monitoring was carried out from a height of 100 meters using the DJI Matrice 600 hexacopter, the Slantrange 3 sensor system. The spectral data were processed using the Slantview software standard for the sensor system, the uncorrected data were calculated in the MathCAD environment. It was found that when using the zenith sensor for atmospheric correction, compliance with the flight regime with respect to the wind direction is critical for the interpretation of the data obtained on the nature of the stresses of vegetation. For the first time for spectral control systems equipped with an anti-aircraft illumination control sensor, it was shown that it is necessary to control the flight mode parameter as an angle of attack because of its significant influence on the obtained spectral data. The authors proposed a method for checking the compliance of the shooting mode by the angle of inclination of the vehicle with the data corrected by the Slantrange system, based on the assessment of the geometry of the images obtained during flight in opposite directions. The methodology proposed by the authors for assessing the stress state of plants is suitable for processing and approximate data if the repeated flight of the UAV is impractical.