天津农学院学报 ›› 2023, Vol. 30 ›› Issue (3): 75-79.doi: 10.19640/j.cnki.jtau.2023.03.014

• 专论与综述 • 上一篇    下一篇

农作物病害图像识别研究进展和存在问题

苏丹1, 邓永卓2,通信作者   

  1. 1.天津农学院 计算机与信息工程学院,天津 300392;
    2.天津市农业发展服务中心,天津 300061
  • 收稿日期:2021-07-09 出版日期:2023-06-30 发布日期:2023-09-06
  • 通讯作者: 邓永卓(1981—),男,高级农艺师,学士,主要从事农业信息化和农业技术推广工作。E-mail:dengyongzhuo@qq.com。
  • 作者简介:苏丹(1997—),女,硕士在读,主要从事农业信息化研究。E-mail:supersdan@163.com。

Research progress and problems in crop disease image recognition

Su Dan1, Deng Yongzhuo2,Corresponding Author   

  1. 1. College of Computer and Information Engineering, Tianjin Agricultural University, Tianjin 300392, China;
    2. Tianjin Agricultural Development Service Center, Tianjin 300061, China
  • Received:2021-07-09 Online:2023-06-30 Published:2023-09-06

摘要: 针对卷积神经网络训练过程中耗费时间长以及大量参数设定等问题,提出采用卷积神经网络结合迁移学习实现智能、快速、准确识别农作物病害类型至关重要。本文首先介绍了农作物病害识别的发展进程,然后介绍了农作物病害识别方法的国内外研究进展,同时分析了其在农作物病害识别上存在的优缺点,最后指出了目前农作物病害识别存在的环境、模型结构和硬件条件等问题,并对农作物病害识别未来的研究方向进行了展望。

关键词: 农作物病害识别, 深度学习, 迁移学习, 研究进展

Abstract: Aiming at the problems of time-consuming and low efficiency of traditional identification methods, it is very important to realize intelligent, fast and accurate identification of crop disease types. This paper first introduces the development process of crop disease identification, then introduces the research progress of crop disease identification methods at home and abroad, and analyzes its advantages and disadvantages in crop disease identification. Finally, it points out the problems of current crop disease identification, such as environment, model structure and hardware conditions, so as to provide reference for future study on crop disease identification.

Key words: crop disease identification, deep learning, transfer learning, research progress

中图分类号: