Plant leaf pest identification method based on improved convolutional neural network

Abstract

The invention discloses a plant leaf pest identification method based on an improved convolutional neural network. According to the method, an improved AlexNet network model is provided, a convolutional neural network model combining batch normalization and global pooling is adopted to recognize various leaf diseases and insect pests, the improved model is greatly optimized in training time and memory requirements, model parameters are simplified, and meanwhile, the generalization ability of the model is also improved; according to the invention, the trained improved AlexNet feature network isused as a plant disease and insect pest identification model, so that the accuracy of plant leaf disease and insect pest identification is improved, the robustness is better, different diseases and insect pests of various plant leaves can be identified, and resources and time required by model training are reduced.

Description

technical field [0001] The invention belongs to the technical field of image recognition, and in particular relates to a method for identifying plant leaf diseases and insect pests based on an improved convolutional neural network. Background technique [0002] In recent years, the global climate has intensified and deteriorated, and the frequency and severity of agricultural meteorological disasters and pests and diseases in my country have increased, which greatly threatens my country's food security. As the ecological environment becomes more and more fragile, the phenomenon of crop diseases and insect pests is becoming more and more serious. Due to the decline in the quality of the plant growth environment, the spread of diseases and insect pests has been accelerated. Only by timely obtaining and identifying information on crop diseases and insect pests can effective control measures be taken. Although the use of chemical pesticides can control Plant diseases, but due to...

AI Technical Summary

Problems solved by technology

[0005] In view of the above-mentioned deficiencies in the prior art, the method for identifying plant leaf diseases and insect pests based on the improved convolutional neural network provided by the present invention solves the problems of long convergence time, large model parameters and low recognition accuracy in the training of existing plant disease and insect pest identification models.

Images