湖北农业科学 ›› 2019, Vol. 58 ›› Issue (16): 129-132.doi: 10.14088/j.cnki.issn0439-8114.2019.16.030

• 信息工程 • 上一篇    下一篇

基于计算机视觉的小麦叶面积测量

郝雅洁, 张吴平, 史维杰, 赵明霞, 吕致, 李富忠   

  1. 山西农业大学软件学院,山西 太谷 030801
  • 收稿日期:2019-06-03 出版日期:2019-08-25 发布日期:2019-11-12
  • 通讯作者: 李富忠。
  • 作者简介:郝雅洁(1994-),女,山西太原人,在读硕士研究生,研究方向为计算机视觉,(电话)15234492439(电子信箱)734423793@qq.com
  • 基金资助:
    农业大数据创新平台项目(K481811076); 山西省重点研发计划(指南)项目(201703D221033-3)

Measurement of wheat leaf area based on computer vision

HAO Ya-jie, ZHANG Wu-ping, SHI Wei-jie, ZHAO Ming-xia, LYU Zhi, LI Fu-zhong   

  1. College of Software,Shanxi Agricultural University,Taigu 030801,Shanxi,China
  • Received:2019-06-03 Online:2019-08-25 Published:2019-11-12

摘要: 运用计算机视觉技术对目标作物小麦的侧拍、俯拍图像进行识别处理,计算出相应叶片所占面积大小,从多变量因素对其测量分析,与叶真实面积大小进行比较,分析建立关系模型。结果表明,小麦植株的侧拍、俯拍面积与叶真实面积之间存在回归关系,且相关性较高,R2值达到0.91。且经过验证,测量结果较准确,说明此回归模型可行。

关键词: 计算机视觉, 小麦, 叶面积, 像素数

Abstract: The computer vision technology was used to identify the lateral and overhead images of the target crop wheat, and the corresponding leaf area was calculated. The multi-variable factors were used to measure and analyze it. Finally, the real area of the leaf was compared and analyzed to establish a relationship model. The results showed that there was a regression relationship between the lateral beat, the overshoot area and the true leaf area of the wheat plants, the correlation was high, and the R2 value reached 0.91. After verification, the measurement results are more accurate, indicating that this regression model is feasible.

Key words: computer vision, wheat, leaf area, number of pixels

中图分类号: