欢迎访问《中国农学通报》,

中国农学通报 ›› 2023, Vol. 39 ›› Issue (19): 153-158.doi: 10.11924/j.issn.1000-6850.casb2023-0171

• 农业信息·科技教育 • 上一篇    下一篇

近红外叶片水分无损检测系统设计与试验

郭新东(), 孙瑜(), 张永梅   

  1. 山西农业大学信息科学与工程学院,山西太谷 030801
  • 收稿日期:2023-03-03 修回日期:2023-05-12 出版日期:2023-07-05 发布日期:2023-07-03
  • 通讯作者: 孙瑜,女,1988年出生,陕西延安人,讲师,硕士研究生,研究方向为近红外无损检测。通信地址:030801 山西省晋中市太谷区铭贤南路1号 山西农业大学,E-mail:sunyu@sxau.edu.cn
  • 作者简介:

    郭新东,男,1990年出生,山西忻州人,讲师,博士研究生,研究方向为图像处理和无损检测。通信地址:030801 山西省晋中市太谷区铭贤南路1号 山西农业大学,E-mail:

  • 基金资助:
    国家自然科学基金“数字图像处理技术辅助相场法模拟压电织构陶瓷晶粒取向生长”(52102138); 山西省高等学校科技创新项目“基于三维点云的农田道路场景语义分割”(2022L086)

Design and Experiments of Nondestructive Moisture Content Testing System Based on Near-infrared Spectroscopy

GUO Xindong(), SUN Yu(), ZHANG Yongmei   

  1. College of Information Science and Engineering, Shanxi Agriculture University, Taigu, Shanxi 030801
  • Received:2023-03-03 Revised:2023-05-12 Online:2023-07-05 Published:2023-07-03

摘要:

玉米叶片水分含量是玉米生长过程中的重要生理指标,近红外线水分仪能够有效检测玉米叶片含水率。传统近红外水分仪存在功耗大、成本高等不足,为此使用1450 nm的红外二极管作为光源,设计基于STM32的便携式近红外玉米叶片水分无损检测系统。首先设计检测系统,包括光源电路、信号调理电路、STM32最小系统、电源电路等。然后对近红外水分仪进行水分标定,采集叶片吸收光和叶片实际水分含量,建立光吸收率与叶片含水率的线性拟合函数,并将拟合函数写入STM32小型系统。最后对水分检测仪进行测试,测量并实时显示叶片含水率。对玉米叶片的试验结果表明,光吸收率和叶片含水率耦合模型建模集相关系数Rc为0.8019,RMSEC为0.51,测试集最大测量误差1.80%,测量误差平均值为0.964%,可以满足农业生产要求。搭建的系统具有检测速度快、体积小、低成本的优点,可为研究近红外的光谱检测仪器提供理论参考和技术支持。

关键词: 近红外光, 无损检测, 玉米叶片, 含水率, STM32

Abstract:

The moisture content of corn leaves is an important physiological index in the process of crop growth. Near-infrared moisture meter can effectively detect the moisture content of corn leaves. Since the traditional near- infrared moisture meter has the disadvantages of high power consumption and high cost, we design a portable near-infrared crop leaf moisture nondestructive testing system based on STM32 by using a LED light source with 1450 nm. First, we designed the testing system consisting of light source circuit, signal processing circuit, STM32 minimum system, power circuit, etc. Then the calibration of the system was performed, which included collecting the reflected light and water content of the leaves, establishing linear fitting model of leaf water content and light segment. Finally, the portable moisture detector was tested to measure and display the leaf moisture content in real time. The results shows that the correlation coefficient of leaf moisture content and light absorption of the portable moisture detector is 0.8019 with RMSEC of 0.51, and the maximum measurement error of the test set is 1.80%, and the average measurement error is 0.964%, which meet the need of agricultural industry. This system has the advantages of fast detection speed, small size and low cost, which provides theoretical reference and technical support for the study of near infrared spectrum detection instruments.

Key words: near-infrared spectroscopy, nondestructive testing, maize leaves, moisture content, STM32