基于激光散射图像小麦叶片叶绿素检测研究

Detection of chlorophyll content of wheat leaves based on laser scattering images

为了实现叶绿素含量的无损检测研究,采用激光后向散射图像技术来测量小麦叶片光学特性参量的方法,进行了理论分析和实验验证。利用670nm和970nm的半导体激光器和视频成像系统获得了小麦叶片的激光后向散射图像,通过漫反射理论分析了叶片组织表面的漫射光分布和在这两个波长下绿叶、黄叶、干叶的激光后向散射图像的变化特征,取得了其光学特性参量数据(约化散射系数和吸收系数),并与叶片的叶绿素相对含量值建立对应的函数关系。结果表明,小麦叶片的光学特性参量与叶绿素相对含量值呈现线性相关,其中利用约化散射系数建立的叶绿素相对含量值预测模型中,预测集样本的相关系数为0.9095,预测均方根误差为5.9;利用吸收系数建立的叶绿素相对含量值预测模型中,预测集样本的相关系数为0.8366,预测均方根误差为7.5,说明激光后向散射图像技术测定植物叶绿素含量是可行的。这一结果对激光散射图像实现农作物长势诊断是有帮助的。

In order to realize nondestructive test of the chlorophyll content, the optical characteristic parameters of wheat leaves were measured with laser backscatter images. Firstly, the laser backscatter images of wheat leaves were obtained with 670nm and 970nm of semiconductor lasers and a video imaging system. Secondly, based on the diffuse theory, diffusive light distribution of the leaf surface was analyzed, and the laser backscatter images of green leaves, yellow leaves and dry leaves were compared at both the above wavelengths. The optical characteristic parameter data （the reduced scattering and absorption coefficient） were obtained, and the function corresponding to the relative contents of chlorophyll was established. The results show the optical characteristic parameters of wheat leaves and are linear with the relative contents of chlorophyll. In the model predicting the relative contents of chlorophyll value based on the scattering coefficients, the correlation coefficient prediction samples is set 0. 9095, prediction root mean square error （ RMSE ） is 5.9. In the model predicting the relative contents of chlorophyll value based on the absorption coefficients, the correlation coefficient prediction samples is set 0. 8366, prediction RMSE is 7.5. The results show that it is feasible to detect plant chlorophyll content and diagnose the crop growth conditions with laser backseatter image technology.