Identification of Citrus Greening (HLB) Using a VIS-NIR Spectroscopy Technique

Abstract

Citrus greening, also known as Huanglongbing or HLB, is a major threat to the U.S. citrus industry. Currently, scouting and visual inspection are used for screening infected trees. However, this is a time-consuming and expensive method for HLB disease detection. Moreover, as it is subjective, the current method exhibits high detection error rates. The objective of this study was to investigate the potential of visible and near-infrared (VIS-NIR) spectroscopy for identifying HLB-infected citrus trees. The spectral data from infected and healthy orange trees of the Valencia variety were collected from four different orchards in Florida. Two different spectroradiometers with a spectral range of 350 to 2500 nm were used to collect the canopy reflectance spectral data. Three classification techniques were used to classify the data: k-nearest neighbors (KNN), logistic regression (LR), and support vector machines (SVM). Analysis showed that using only one canopy reflectance observation per tree was inadequate. None of the classification methods was successful in discriminating healthy trees from HLB-infected trees because of the large variability in the canopy reflectance spectral data. When five spectra from the same tree were used for classification, the SVM and weighted KNN methods classified spectra with 3.0% and 6.5% error rates, respectively. The results from this study indicate that canopy VIS-NIR spectral reflectance data can be used to detect HLB-infected citrus trees; however, high classification accuracy (>90%) requires multiple measurements from a single tree.