摘要
为了研究基于图像红(R)、绿(G)、蓝(B)颜色参数和叶片SPAD值预测光合作用指标的可行性,以草莓叶片为试验材料,构建多元线性回归模型和反向传播(BP)神经网络模型,对叶片蒸腾速率、气孔导度、净光合速率、胞间CO_(2)浓度进行估测,并对其精度进行评价和验证。结果表明,基于BP神经网络模型,使用图像RGB颜色参数和SPAD值对叶片蒸腾速率进行预测的效果较好,其次是气孔导度。BP神经网络模型的估测精度高于多元线性回归模型,蒸腾速率、气孔导度、净光合速率和胞间CO_(2)浓度的模型预测准确率分别达到91.5%、83.3%、74.4%和71.5%。BP神经网络的蒸腾速率模型、气孔导度模型的决定系数(R2)分别为0.9222、0.8423,均方根误差(RMSE)分别为0.0002、0.0259,平均绝对误差(MAE)分别为0.0001、0.0006。由结果可知,通过数码相机采集图像,并构建RGB模型,可简易快速估测草莓叶片蒸腾速率、气孔导度,能用于生产中草莓光合指标的估测。
In order to explore the feasibility of using RGB image feature and SPAD value in photosynthetic indexes prediction,strawberry leaves were selected as experimental materials in this study.Multiple linear regression model and back propagation(BP)neural network model were constructed to estimate leaf transpiration rate,stomatal conductance,net photosynthetic rate and intercellular CO_(2)concentration,and their accuracy was evaluated and verified.The results showed that the prediction of leaf transpiration rate by using RGB color parameters and SPAD values based on BP neural network model was better,followed by stomatal conductance.The estimation accuracy of BP neural network model was higher than that of multiple linear regression model,and the prediction accuracy of transpiration rate,stomatal conductance,net photosynthetic rate and intercellular CO_(2)concentration reached 91.5%,83.3%,74.4%and 71.5%,respectively.The determination coefficients(R^(2))of transpiration rate model and stomatal conductance model based on BP neural network were 0.9222 and 0.8423,the root mean square errors(RMSE)were 0.0002 and 0.0259,and the mean absolute errors(MAE)were 0.0001 and 0.0006,respectively.Therefore,the transpiration rate and stomatal conductance of strawberry leaves can be easily and quickly estimated by using digital camera to collect images and construct RGB model,which can be used to predict photosynthetic indexes of strawberry in production.
作者
樊小雪
李德翠
李远
任妮
FAN Xiao-xue;LI De-cui;LI Yuan;REN Ni(Institute of Agricultural Information,Jiangsu Academy of Agricultural Sciences,Nanjing 210014,China;Key Laboratory of Smart Agricultural Technology(Yangtze River Delta),Ministry of Agriculture and Rural Affairs,Nanjing 210014,China)
出处
《江苏农业学报》
CSCD
北大核心
2024年第4期675-681,共7页
Jiangsu Journal of Agricultural Sciences
基金
江苏省农业科技自主创新基金项目[CX(22)5007]。
