摘要
为研究利用手机图像预测玉米光响应曲线特征参数的可行性,通过自主设计的便携式图像采集装置,获取不同施氮量下滴灌玉米大喇叭口期冠层图像,提取其冠层图像特征参数,计算玉米光响应特征参数表观量子效率(α)、最大净光合速率(Pnmax)、光补偿点(LCP)和暗呼吸速率(Rd),找出与光响应特征参数相关性高的归一化冠层覆盖系数(CC)作为自变量,建立CC与玉米光响应曲线特征参数间动态模型,并根据模型评价指标R^2、RMSE和nRMSE筛选出各参数的最优模型。结果表明,CC与α的最优模型为有理函数模型,与Pnmax最优模型为幂函数模型,与LCP最优模型为指数函数模型,与Rd以二次多项式模型为最优;各反演模型的R2均不小于0.876,RMSE介于0.002~3.673μmol/(m^2·s)之间,nRMSE不超过9.071%,且各模型验证集的R^2均不小于0.833,RMSE均不大于5.989μmol/(m^2·s),nRMSE不超过9.659%。将数字图像特征参数与作物光响应曲线特征参数有机结合,可为玉米光响应曲线特征参数的快速获取提供一定的理论依据。
The characteristic parameters of light response can indicate the process of photosynthesis,capacity of photosynthesis,and response of adversity stress in crops.In order to explore the feasibility of using cellphone photos to predict the light response characteristic parameters of maize,canopy images at the big flare stage of the drip irrigation maize under different nitrogen levels were obtained by the self-developed portable image acquisition device.Feature parameters were extracted from the canopy image,and the photosynthetic physiological characteristics parameters were calculated,such as apparent quantum efficiency(α),dark respiration rate(R d),light compensation point(LCP)and maximum net photosynthesis rate(P nmax).A normalized canopy cover factor(C C)was highly correlated with the light response characteristic parameters,as an independent variable was used to predict these parameters.The optimal model was selected according to the model evaluation indicators such as R^2,RMSE and nRMSE.The results showed that the optimal model of C C andαwas the rational function model,the optimal model of P nmax was the power function model,the optimal model of LCP was the exponential function model and the optimal model of R d was the quadratic polynomial model.The R 2 values of each model were greater than 0.876,the values of RMSE were between 0.002μmol/(m^2·s)and 3.673μmol/(m^2·s),and the nRMSE was no more than 9.071%.Meanwhile,the R^2 values of each model validation set were greater than 0.833,the RMSE values were less than 5.989μmol/(m^2·s),and the nRMSE values were no more than 9.659%.Combining the digital image feature parameters with the maize light response curve characteristic parameters method,it was recommended to quickly acquire characteristic parameters of maize light response curve,which provided a theoretical basis for the light response.
作者
贾彪
贺正
JIA Biao;HE Zheng(School of Agriculture,Ningxia University,Yinchuan 750021,China)
出处
《农业机械学报》
EI
CAS
CSCD
北大核心
2019年第7期229-236,共8页
Transactions of the Chinese Society for Agricultural Machinery
基金
国家自然科学基金项目(31560339)
宁夏高等学校科研项目(NGY2017025)
宁夏回族自治区科技重大专项(2018BBF0200404)
宁夏区重点研发计划项目(2018BBF02018)
“十二五”国家科技支撑计划项目(2015BAD22B01)
关键词
玉米
手机图像
光响应曲线参数
归一化
冠层覆盖系数
maize
cellphone photos
light response curve parameters
normalized
canopy cover coefficient
