基于卷积神经网络的植物病虫害识别研究综述

Review on Plant Disease and Pest Identification Based on Convolutional Neural Network

植物病虫害严重影响植物的生长与生产,对其进行及时精准的识别与管控,能有效提升植物的产量和质量。近年来,深度学习发展迅猛,卷积神经网络作为深度学习的代表算法之一,具有较好的图像分类和识别能力,广泛应用于植物病虫害的识别研究。对近几年基于卷积神经网络的植物病虫害识别研究进行综述;简要介绍几种基础网络的模型结构、网络结构优化方法、卷积神经网络与其他方法的结合应用等;探讨目前基于卷积神经网络的植物病虫害识别研究的热点难点,并对其应用前景进行展望。

Plant diseases and insect pests have seriously affected the growth andproduction of plants.Timely accurate identification and controlof plant diseases andinsect pests can effectively improve the yield and quality of crops.In recent years,deep learning methods have developed rapidly.As one of the representative algorithms of deep learning,convolutional neural network has excellent image classification and recognition capabilities,and has been widely used in the identification of plant diseases and insect pests.The research on plant disease and pest identification based on convolutional neural network is reviewed.The structure and characteristics of several basic network models,the optimization methods of network structure and the combined application of convolutional neural network and other methods are reviewed;the difficulties of convolutional neural network based on plant disease and pest identification are discussed,and its application prospect is prospected.