Crop disease prediction method based on crop ontology concept response

Abstract

The invention discloses a crop disease prediction method based on crop ontology concept response. The method comprises the following steps: defining a local ontology sub-concept, modeling a disease ontology, constructing a deep learning model, detecting the ontology concept, predicting a disease result based on supervised learning, and predicting the disease result based on concept reasoning. According to the method disclosed in the invention, based on establishment of disease knowledge ontology, a corresponding ontology concept detector is trained based on the deep learning model; then, afterconcept response is detected, disease result prediction is carried out in combination with supervised learning and concept reasoning technologies; on one hand, when disease result prediction is carried out based on supervised learning, the noise in the label is attempted to be filtered; on the other hand, the uncertainty of concept response is reduced through concept reasoning, the relationship between crop state expression and disease results is fully analyzed, disease inversion is promoted while the disease prediction reliability is improved, and the problem of insufficient marked data is prevented.

Description

technical field [0001] The invention relates to the technical field of crop disease prediction, in particular to a crop disease prediction method based on the crop ontology concept response. Background technique [0002] As the lifeline of the national economy, agriculture is an important foundation for my country's economic construction and development, and crops are an important resource for people's lives. However, crops are often infected with various diseases due to various stresses during the growth process, and the occurrence of diseases will affect The healthy growth of crops interferes with the important physiological functions of crops, which in turn leads to damage to crop quality and yield decline, which brings incalculable losses to agricultural production. Especially in recent years, the planting area, variety and quantity of crops in my country have gradually increased, but Problems in climate, ecological environment and planting system lead to frequent crop dis...

AI Technical Summary

Problems solved by technology

[0002] As the lifeline of the national economy, agriculture is an important foundation for my country's economic construction and development, and crops are an important resource for people's lives. However, crops are often infected with various diseases due to various stresses during the growth process, and the occurrence of diseases will affect The healthy growth of crops interferes with the important physiological functions of crops, which in turn leads to damage to crop quality and yield decline, which brings incalculable losses to agricultural production. Especially in recent years, the planting area, variety and quantity of crops in my country have gradually increased, but Problems in climate, ecological environment and planting system lead to frequent crop diseases, which have a serious negative impact on agricultural production and even the entire national economy. According to statistics, the area of ​​crops affected by diseases is as high as billions of mu every year, directly food Both crop losses and indirect economic crop losses are tens of billions of catties, and these problems are still showing a trend of worsening year by year, seriously threatening the healthy development of agriculture;

[0003] Traditional crop disease detection usually uses chemical reagents for off-line analysis and testing in the laboratory. This method is complex and pollutes the environment. At the same time, with the rapid development of imaging technology and image processing technology, researchers have begun to use computer vision The monitoring of crop disease characteristics by various means, such as hyperspectral image technology, and the existing research on crop disease detection based on hyperspectral images often only analyze single parts such as leaves or fruits of crops, ignoring the fact that diseases will occur in crops at different stages of development. The fact that each part is reflected, and the disease prediction based on the end-to-end deep learning model also lacks the analysis and inversion of the disease results. Therefore, the present invention proposes a crop disease prediction method based on the crop ontology concept response to solve the existing problems. problems in technology

Images