Problems and Prospects of Using Artificial Intelligence to Monitor Phytosanitary Conditions of Crops

Abstract

Recently, there was an increase in the incidence of crops by diseases. It led to crop shortages of up to 30% or more. An urgent task is the timely detection of diseases to take measures on plant protection and prevent the spread of harmful organisms. Due to the complexity and large scale of the ongoing work to identify and account for diseases of crops, the use of modern digital technologies and AI in this area looks promising. For this purpose, most often, algorithms of the neural network are used, namely, the convolutional neural network algorithm, which has multi-layer and most accurate output information. However, the use of these technologies is also fraught with several difficulties, in particular with the differentiated diagnosis of infectious and non-infectious plant diseases that require different protective measures, the inability to diagnose diseases of the plant root system, and the high cost of creating a database of affected plants. However, these problems can be solved. AI has great prospects in monitoring the phytosanitary state of crops.