Deep learning-based pest control scheme generation method and system, and storage medium

Abstract

The invention discloses a pest control scheme generation method and system based on deep learning and a storage medium, and relates to the technical field of pest control, and the pest control scheme generation method comprises the steps: obtaining plant image information in a target area, according to the plant image information, judging whether the plants in the target area have insect damage conditions or not, determining current insect damage symptoms, performing representation in a heterogeneous graph form based on insect damage types, the insect damage symptoms and a prevention and control method, and constructing a prevention and control scheme generation model based on a graph convolutional neural network; and inputting the current insect pest symptom as an initial node into a prevention and control scheme generation model to obtain a prediction score of the prevention and control scheme, and taking the prevention and control scheme with the high prediction score as an insect pest prevention and control scheme suitable for the plants in the target area. According to the method, the potential relationship between the insect pest and the control scheme is obtained through the graph convolutional neural network, so that the pertinence and accuracy of the control scheme are greatly improved, and meanwhile, the formulating efficiency of the plant insect pest control scheme is improved.

Description

technical field [0001] The invention relates to the technical field of pest control, and more particularly, to a method, system and storage medium for generating a pest control scheme based on deep learning. Background technique [0002] The traditional extensive agricultural management method not only fails to greatly promote the efficiency of agricultural production, but also greatly wastes agricultural resources and causes damage to the ecological environment. With the development of information technology, precision agricultural production methods have emerged as the times require, especially the prevention and management of pests and diseases has shown a trend of intelligence; the current traditional methods of identification and control of plant pests mainly rely on field inspections for on-site diagnosis of pest species However, such methods have low efficiency and poor timeliness, and are far from meeting the needs of my country's agricultural development. However, d...

AI Technical Summary

Problems solved by technology

[0002] The traditional extensive agricultural management method not only fails to greatly promote the efficiency of agricultural production, but also greatly wastes agricultural resources and causes damage to the ecological environment

With the development of information technology, precision agricultural production methods have emerged as the times require, especially the trend of intelligent pest control and management; at present, traditional plant pest identification and control methods mainly rely on field inspections for on-site diagnosis of pest types , and this kind of method is inefficient and time-sensitive, which is far from meeting the needs of my country's agricultural development.

However, due to the limited ability of scientific and technological knowledge, agricultural personnel cannot timely grasp the occurrence and development of plant diseases and insect pests, and often miss the best control period. Seriously pollute the environment, therefore, it is particularly important to formulate a scientific and reasonable control plan according to the pest situation

Images