Development of Real-Time Monitoring and Early Warning System Platform of Internet of Things for Potato Late Blight

  • Yanping Chen
Keywords: Potato Late Blight, Early Warning System Prediction, Agricultural Internet of Things, Feature Extraction, Intelligent Recognition


Although the planting area of potato in China is large, its yield and quality are affected by various diseases and
insect pests. This paper aims to establish a real-time monitoring and early warning system platform for potato
late blight based on the number of Phytophthora spores in the air. Firstly, the common detection methods of
potato late blight and related technologies of Internet of things were analyzed. Secondly, Phytophthora spores
are the internal cause of the outbreak of potato late blight. Monitoring the number of spores and cooperating
with the meteorological conditions of the outbreak can achieve a more accurate broadcast of the epidemic.
Using machine vision technology to complete spore automatic identification and counting has become the key
technology of early warning system. The experimental results show that the degree of spore diffusion in the
disease prediction module is determined by the results of identification and counting. When the spore number is
greater than 20, the diffusion condition input is 2; when the spore number is between 2 and 20, the diffusion
condition input is 1; when the spore number is less than or equal to 2, the diffusion condition input is 0. The
input values of other conditions are the actual results measured by the meteorological department or relevant