欢迎访问《中国农学通报》，
中国农学通报

中国农学通报 ›› 2020, Vol. 36 ›› Issue (16): 149-155.doi: 10.11924/j.issn.1000-6850.casb20191100832

所属专题： 水稻

• 农业信息·科技教育 • 上一篇    下一篇

计算机视觉技术在水稻氮素营养诊断中应用的研究进展

杨红云1,*(), 罗建军2, 万颖2, 孙爱珍2   

  1. 1.江西农业大学软件学院/江西省高等学校农业信息技术重点实验室,南昌 330045
    2.江西农业大学计算机与信息工程学院,南昌 330045
  • 收稿日期:2019-11-14 修回日期:2020-02-23 出版日期:2020-06-05 发布日期:2020-05-20
  • 通讯作者: 杨红云
  • 基金资助:
    国家自然科学基金项目“基于机器学习的水稻生长过程建模方法研究”(61562039);国家自然科学基金项目“物理化学交互式的水稻叶片细胞可视化建模研究”(61762048);国家自然科学基金项目“基于生理生态的水稻根系建模及可视化仿真研究”(61862032);江西省教育厅科技项目“水稻生长信息的计算机视觉获取方法研究”(GJJ160374);江西省教育厅科技项目“基于视觉技术的水稻营养智能诊断与建模方法研究”(GJJ170279)

Application of Computer Vision Technology in Nitrogen Nutrition Diagnosis of Rice: Research Progress

Yang Hongyun1,*(), Luo Jianjun2, Wan Ying2, Sun Aizhen2   

  1. 4.School of Software, Jiangxi Agricultural University/ Key Laboratory of Agricultural Information Technology of Colleges and Universities in Jiangxi Province, Nanchang 330045
    2.School of Computer and Information Engineering, Jiangxi Agricultural University, Nanchang 330045
  • Received:2019-11-14 Revised:2020-02-23 Online:2020-06-05 Published:2020-05-20
  • Contact: Yang Hongyun

RichHTML

2

PDF (PC)

74

摘要：

为探究水稻冠层外观特征与水稻氮素营养状况的关系,综述了应用计算机视觉技术进行水稻氮素营养诊断的基本思路及研究进展。主要分述了常规性水稻氮素营养指标的获取,水稻冠层图像的获取、预处理以及特征的提取与优化,水稻氮素营养诊断模型建立的方法等方面的内容。指出近地面水稻冠层图像的获取方法、水稻图像处理方法、多种综合性特征的优化选择方式都有待进一步研究,应用机器学习进行水稻氮素营养诊断建模的方法需要更加深入探究。今后应加大计算机视觉技术在水稻氮素营养诊断相关研究中的应用,将多种图像处理方法、特征优化选择方法与机器学习建模方法相结合,并探寻更为简便易行的方法进行水稻氮素营养诊断。

关键词: 水稻, 图像处理, 特征提取与优化, 计算机视觉技术, 机器学习, 氮素营养诊断

Abstract:

To explore the relationship between the appearance characteristics of rice canopy and the nitrogen nutrition status of rice, the basic ideas and research progress of the application of computer vision technology in rice nitrogen nutrition diagnosis were summarized. This paper mainly discusses the acquisition of conventional rice nitrogen nutrition index, the acquisition and preprocessing of rice canopy image, the extraction and optimization of features, and the establishment of rice nitrogen nutrition diagnosis model. It is pointed out that the methods of rice canopy image acquisition, rice image processing and optimization of multiple comprehensive characteristics need to be further studied, and the methods of using machine learning to diagnose and model rice nitrogen nutrition need to be further explored. In the future, it is necessary to increase the application of computer vision technology in the research of rice nitrogen nutrition diagnosis, and combine multiple image processing methods, characteristics optimization selection methods and machine learning modeling methods together to explore a more simple and easy method for rice nitrogen nutrition diagnosis.

Key words: rice, image processing, characteristics extraction and optimization, computer vision technology, machine learning, nitrogen nutrition diagnosis

中图分类号: