Plant disease detection method based on residual network

Abstract

The invention relates to a plant disease detection method based on a residual network, and the method comprises the following steps: building a plant disease detection data set which adopts a brand new leaf disease image data set of AI Challenger; carrying out statistical analysis on the categories and the number of the data sets, separating each type of images from the complete data set, storingthe images in corresponding category folders, and then carrying out image preprocessing; performing enhancement processing on the image by using random angle rotation, random horizontal mirror rotation, random vertical mirror rotation and random affine transformation methods; setting hyper-parameters of deep convolutional neural network training; using pyTorch as a deep learning framework, using aresidual network as a deep learning model, and training the residual network deep learning model to obtain a plant disease detection result. The residual error network ResNet is used for training thedeep convolutional neural network, 27 diseases of up to 10 crops can be accurately identified, the speed is high, the precision is high, and the method has important significance for agricultural production.

Description

technical field [0001] The invention belongs to the technical field of image recognition, in particular to a plant disease detection method based on a residual network ResNet. Background technique [0002] Plants play a vital role in human production, life and even survival. However, due to the continuous expansion of human production and life, excessive hunting and gathering behavior of humans, pollution from many factories, and the development of large-scale land vegetation, the natural ecosystem has been greatly changed, making the environmental adaptability of plants lower and self-regulation disorder , causing plant diseases. [0003] Plant diseases are mainly caused by biological factors (biological pathogens) and abiotic factors (unsuitable living environment), which make plants grow abnormally and can cause plants to die within a certain period of time. Long-term illness can lead to the extinction of plant species or infect other species to extinction. This effect...

AI Technical Summary

Problems solved by technology

[0004] For many years, the diagnosis of plant diseases has been mainly based on the experience of plant disease workers, which has the disadvantages of high work intensity, the influence of subjective factors, and the dependence on agricultural experts

In recent years, the use of artificial neural networks and deep learning to detect plant diseases has become a new trend. After searching, it is found that the existing literatures on plant disease detection by artificial neural networks or deep learning are all focused on plant disease detection from different emphases. There are disadvantages such as low accuracy and slow recognition speed

Images