基于计算机视觉的水稻叶绿素含量测定

Determination of Chlorophyll Content in Rice Based on Computer Vision

叶绿素含量影响植物的各种生理机能,准确测定水稻叶绿素含量有利于了解水稻的光合作用能力,为高产育种和栽培提供依据。叶绿素含量的测定方法有多种,传统测定方法步骤繁琐且耗时费力,而便携式叶绿素仪只能进行点测定。计算机视觉是一种快速便捷的图像处理技术,可以用于作物的色素含量测定和营养状况诊断。为此,设计了一种基于计算机视觉的水稻叶绿素含量测定方法,叶片图像通过扫描获得,经过处理并提取叶片轮廓后以G-R、B-R和R/(G+B)等3个颜色特征参数建立叶绿素含量的估算模型。结果表明:以G-R和R/(G+B)建立的模型精度较高,对检验样本SPAD值拟合方程的决定系数和斜率都接近1,可以用来准确测定水稻的叶绿素含量。

Chlorophyll content affects various physiological functions of plants.Accurate determination of chlorophyll content in rice is beneficial to understanding the photosynthesis ability of rice,and provides a basis for high-yield breeding and cultivation.There are many methods for determining the chlorophyll content.The traditional method of measurement is cumbersome and time consuming,and the portable chlorophyll meter can only perform point determination.Computer vision is a fast and convenient image processing technology that can be used for the determination of pigment content and nutritional status of crops.In this paper,a method for determining the chlorophyll content of rice based on computer vision was designed.The image of the leaf was obtained by scanning.After processing and extracting the contour of the leaf,the chlorophyll content was established by the three color characteristic parameters of GR,BR and R/(G+B).Estimate the model.The model established by G-R and R/(G+B)has high precision.The coefficient of determination and slope of the fitting equation for the SPAD value of the test sample are close to 1,which can be used to accurately determine the chlorophyll content of rice.