Monitoring of agricultural land productivity using unmanned aerial vehicles and artificial neural networks

Abstract

Abstract The results of the analytical review of the use of unmanned aerial vehicles (UAVs) and artificial neural networks in agricultural production are presented. They can be used as aerial robots that perform the function of aerial photography, transportation of technological components, such as plant protection products and perform other similar functions. On the aircraft, some other functional equipment can be installed: thermal imagers, multispectral and IR cameras, etc. With the help of the data obtained from the UAV, it is possible to create an orthophotoplan or 3D model of the terrain, create a map of heights, determine the state of the field, crops and determine their vegetation indices NDVI. The multi-level classification of areas of application of UAVs in agricultural production is proposed. Classification involves the ordering of areas of application of UAV in agriculture depending on the composition in use. A conceptual model of a software package designed to obtain and process remote sensing data using UAVs in different parts of the spectrum has been developed. The software package is designed to obtain and process the results of monitoring and subsequent analysis of the totality of the calculated vegetation indices. The main research tasks solved by the developed software, which determine its structure, are formulated. To predict the yield of different crops, a method of applying the results of aerial photography in conjunction with experimental data on the biological development of crops has been developed. For the practical use of the developed methodology, a database for each culture is formed. The obtained results are used to construct regression and matrix mathematical models of the relationship of optical-spectral characteristics with the productivity of crops.