MODELS AND METHODS FOR PROCESSING AGRICULTURAL DATA

Abstract

Аn overview of models used in processing video information from small aircraft related to the study of the state of agricultural lands is given. The main ones are: vegetation indices for predicting yield, machine learning and neural networks for detecting diseases and the quality of crop germination, yield forecasting models based on biophysical, statistical, polynomial and fuzzy methods. Biophysical plant growth models simulate the processes of growth and development of crops. Statistical models use historical yield data in combination with various predictors, machine learning algorithms are used to identify patterns and create predictive models. Polynomial models are used to build a yield forecast, fuzzy cognitive maps allow taking into account expert knowledge and describe the degree of mutual influence of factors.