Greenhouse winter jujube disease prediction model based on agricultural Internet of Things and depth belief network

Abstract

The invention provides a greenhouse winter jujube disease prediction model based on the agricultural Internet of Things and a depth belief network. Winter jujube growth environment information is collected and preprocessed. Winter jujube disease video color images are collected, and the features are extracted to form a winter jujube disease joint eigenvector. According to the greenhouse winter jujube disease occurrence regularity, the historical data of winter jujube diseases are combined to establish a greenhouse winter jujube disease joint eigenvector database. The winter jujube disease prediction model based on the depth belief network (DBN) is constructed. The prediction model consists of an input layer, three restricted Boltzmann machines (RBM) and a BP network. An unsupervised greedy layer-by-layer method and the joint eigenvector in the database are used to pre-train the prediction model to acquire the best parameter value of the prediction model. The prediction model is tested to repeatedly train the prediction model. Winter jujube disease prediction is carried out to improve the accuracy of greenhouse winter jujube disease prediction.

Description

technical field [0001] The invention relates to the field of crop disease prediction and machine learning technology, in particular to a greenhouse winter jujube disease prediction model based on the agricultural internet of things and deep belief network. Background technique [0002] The 200,000-acre Dapeng winter jujube in the Yellow River beach area of ​​Dali County, Shaanxi Province, has thin skin, crisp meat, sweet fragrance, and rich nutrition. It is loved by the people all over the country, but there are excessive pesticide residues. Since Dali winter jujube is planted in the Dapeng greenhouse, the high temperature, high humidity and weak light in Dapeng provide suitable conditions for the occurrence of winter jujube diseases, resulting in more types of winter jujube diseases and frequent occurrence of diseases. Over the years, due to the lack of real-time disease information and scientific prevention and control guidance, most fruit farmers regularly spray more than...

AI Technical Summary

Problems solved by technology

However, due to the complexity and diversity of factors for the occurrence of winter jujube diseases in greenhouses, the existing methods and technologies for predicting diseases of winter jujube in greenhouses only use a small amount of environmental information generated by winter jujube and video color images of diseased winter jujube. The accuracy of disease prediction is not high and cannot meet the actual requirements. , there is an urgent need for a method that can accurately predict the disease of winter jujube in greenhouses

Images