Prediction of poppy thebaine alkaloid concentration using UAS remote sensing

被引:0
作者
Faheem Iqbal
Arko Lucieer
Karen Barry
机构
[1] University of Tasmania,School of Technology, Environments and Design
[2] University of Tasmania,Tasmanian Institute of Agriculture
来源
Precision Agriculture | 2020年 / 21卷
关键词
Poppy; Alkaloids; Thebaine; UAS remote sensing; Random forest;
D O I
暂无
中图分类号
学科分类号
摘要
Alkaloid concentration, which represents the quality of industrial poppy, needs to be estimated in a spatially explicit manner to predict the value of crop prior to harvesting. Current practice is to estimate alkaloid concentration using destructive sampling and laboratory analysis. However, in order to estimate the value of the whole crop, a method that could predict alkaloid concentration in field conditions prior to harvesting is needed. In this study, an unmanned aerial system (UAS) with multispectral imaging was tested for estimation of alkaloid concentration of a poppy crop before harvest, which was sown for pharmaceutical purposes in Tasmania, Australia. This study presents the result of a random forest (RF) regression analysis to evaluate the contribution and predictive ability of spectral and structural variables derived from the images. It was found that UAS imagery with an RF model has the potential to estimate thebaine (paramorphine) concentration well before harvesting and without laboratory analysis. It was found that an RF model with the combination of MSAVI, mSR, OSAVI, NDVI and EVI spectral indices can provide optimal results to estimate thebaine with a relative error of 13.56% to 22.36% with training and validation datasets, respectively. The thebaine concentration predicted using the proposed RF model was strongly correlated to the laboratory-measured thebaine concentration, with an R2 value ranging from 0.63 to 0.82 for the training and validation datasets, respectively. These results indicate that poppy thebaine concentration can be estimated with reasonable accuracy 3 weeks prior to harvesting.
引用
收藏
页码:1045 / 1056
页数:11
相关论文
共 129 条
[1]  
Abdel-Rahman EM(2013)Random forest regression and spectral band selection for estimating sugarcane leaf nitrogen concentration using EO-1 Hyperion hyperspectral data International Journal of Remote Sensing 34 712-728
[2]  
Ahmed FB(2014)Land-use and cover classification in a heterogeneous coastal landscape using RapidEye imagery: Evaluating the performance of random forest and support vector machines classifiers International Journal of Remote Sensing 35 3440-3458
[3]  
Ismail R(1981)The effect of environmental factors on growth, development and alkaloid production of poppy. II. Interaction of light and temperature Biochemical Phusiology Pflanzen 176 599-605
[4]  
Adam E(2001)Random forests Machine Learning 45 5-32
[5]  
Mutanga O(1997)On the relation between NDVI, fractional vegetation cover, and leaf area index Remote Sensing of Environment 62 241-252
[6]  
Odindi J(1996)Retrieving leaf area index of boreal conifer forests using Landsat TM images Remote Sensing of Environment 55 153-162
[7]  
Abdel-Rahman EM(2014)Nitrogen status assessment for variable rate fertilization in maize through hyperspectral imagery Remote Sensing 6 6549-6565
[8]  
Bernath J(2002)Derivation of the red edge index using the MERIS standard band setting International Journal of Remote Sensing 23 3169-3184
[9]  
Tetrnyi P(1996)Use of a green channel in remote sensing of global vegetation from EOS-MODIS Remote Sensing of Environment 58 289-298
[10]  
Breiman L(2002)Novel algorithms for remote estimation of vegetation fraction Remote Sensing of Environment 80 76-87