Automatic insect image identification method based on depth convolutional neural network

Abstract

The invention discloses an automatic insect image identification method based on a depth convolutional neural network. The method comprises the following steps: (1), collecting an original image and carrying out pretreatment to form a training set, and calculating a mean value image of the training set; (2), constructing a depth convolutional neural network; (3), collecting a sub image block randomly from a training sample of the training set and carrying out pre training on the depth convolutional neural network by using the sub image block; (4), training the depth convolutional neural network by using the training set and combining a mini-batch-based random gradient descent algorithm; and (5), carrying out pretreatment on a to-be-measured insect image to form a test sample, and using the trained depth convolutional neural network to identify the test sample after subtracting the mean value image of the training set from the test sample. Therefore, the identification precision is high; the identification types are diversified; the insect within-class change robustness is enhanced; and the insect inter-class similarity sensitivity is high.

Description

technical field [0001] The invention relates to the technical field of precision agriculture, in particular to a method for automatic identification of pest images based on a deep convolutional neural network. Background technique [0002] Rice is one of the important food crops in my country. During the whole growth period of rice, there are many diseases, insects and other harmful organisms, especially rice pests, which cause staggering losses every year, directly endangering rice production. At present, my country's rice pest control has always adhered to the "Integrated Pest Management (IPM)" plant protection policy, based on monitoring and forecasting, comprehensively applying agricultural, biological, physical control and chemical control and other technical measures to effectively control pest damage. [0003] The investigation of the types and quantities of rice pests is a basic and important task in the pest forecasting work. If there is no correct survey data, it ...

AI Technical Summary

Problems solved by technology

[0006] However, the existing pest image automatic recognition methods and systems have low recognition rate and poor robustness, and most of them are difficult to push into practical application

Images