河南农业科学 ›› 2023, Vol. 52 ›› Issue (12): 162-171.DOI: 10.15933/j.cnki.1004-3268.2023.12.018

所属专题: 病虫害识别专题

• 农业信息与工程·农产品加工 • 上一篇    下一篇

基于迁移学习和残差网络的谷子病害识别研究

张红涛1,2,罗一铭1,谭联1,杨加蓬1,王宇1   

  1. (1.华北水利水电大学电气工程学院,河南 郑州 450045;2.河南省智慧农业光谱成像检测装备工程技术研究中心,河南 郑州 450045)
  • 收稿日期:2023-07-19 出版日期:2023-12-15 发布日期:2024-01-05
  • 作者简介:张红涛(1977-),男,河南南阳人,教授,博士,主要从事农业信息化、图像处理等研究。E-mail:39583633@qq.com
  • 基金资助:
    国家自然科学基金项目(31671580);河南省重点研发与推广专项(232102110265)

Research on Millet Disease Identification Based on Transfer Learning and Residual Network

ZHANG Hongtao1,2,LUO Yiming1,TAN Lian1,YANG Jiapeng1,WANG Yu1   

  1. (1.College of Electrical Engineering,North China University of Water Resources and Electric Power,Zhengzhou 450045,China;2.Henan Province Intelligent Agriculture Spectral Imaging Testing Equipment Engineering Technology Research
    Center,Zhengzhou 450045,China)
  • Received:2023-07-19 Published:2023-12-15 Online:2024-01-05

摘要: 针对谷子病害提出了一种基于迁移学习和残差网络(Residual CNN)的谷子病害图像识别方法。首先,建立由谷子白发病、谷瘟病、红叶病、锈病等4种病害图像以及正常谷子叶片图像组成的原始样本集;然后,采用基于超绿特征的最大类间方差法实现原始图像的分割,建立谷子病害分割图像数据集,并对该数据集进行扩充处理;最后,基于扩充后的谷子病害分割图像数据集,利用迁移学习和残差网络的思想建立谷子病害识别模型。结果表明,该模型的识别率达到98.2%,相较于基于支持向量机(SVM)的谷子病害识别模型提高了8.9百分点,同时该模型的训练时间相较于基于卷积神经网络(CNN)的谷子病害识别模型减少17.69 min。表明基于迁移学习和残差网络的谷子病害识别模型可有效地对4种谷子叶片病害进行识别。

关键词: 谷子, 病害识别, 图像处理, 计算机视觉, 迁移学习, 残差网络

Abstract: A method of millet disease image recognition based on transfer learning and residual network(Residual CNN)was proposed for millet disease. First,the original sample set was established,whichwas composed of four kinds of disease images including millet white disease,blast,red leaf disease,rustdisease and normal millet leaf image.Then,the original image was segmented by using the maximum inter‐class variance method based on super green feature,the millet disease segmentation image dataset was established,and the dataset was extended. Finally,based on the expanded segmentation image data set of millet disease,the recognition model of millet disease was established by using the idea of transfer learning and residual network. The results showed that the recognition rate of this model reached 98.2%,which was 8.9 percentage points higher than that of the support vector machine(SVM)based millet disease recognition model,and the training time of this model was reduced by 17.69 min compared with that of the convolutional neural network(CNN) based millet disease recognition model.The results indicated that the recognition model of millet disease based on transfer learning and residual network could effectively identify the four kinds of millet leaf diseases.

Key words: Millet, Disease identification, Image processing, Computer vision, Transfer learning, Residual network

中图分类号: