This study demonstrates the applicability of visible-near infrared and thermal imaging for detection of Huanglongbing (HLB) disease in citrus trees. Visible-near infrared (440–900 nm) and thermal infrared spectral reflectance data were collected from individual healthy and HLB-infected trees. Data analysis revealed that the average reflectance values of the healthy trees in the visible region were lower than those in the near infrared region, while the opposite was the case for HLB-infected trees. Moreover, 560 nm, 710 nm, and thermal band showed maximum class separability between healthy and HLB-infected groups among the evaluated visible-infrared bands. Similarly, analysis of several vegetation indices indicated that the normalized difference vegetation index (NDVI), Vogelmann red-edge index (VOG) and modified red-edge simple ratio (mSR) demonstrated good class separability between the two groups. Classification studies using average spectral reflectance values from the visible, near infrared, and thermal bands (13 spectral features) as input features indicated that an average overall classification accuracy of about 87%, with 89% specificity and 85% sensitivity could be achieved with classification models such as support vector machine for trees with symptomatic leaves.

Laurel wilt is a lethal disease of plants in the Lauraceae plant family, including avocado (Persea americana). This devastating disease has spread rapidly along the southeastern seaboard of the United States and has begun to affect commercial avocado production in Florida. The main objective of this study was to evaluate the potential to discriminate laurel wilt-affected avocado trees using aerial images taken with a modified camera during helicopter surveys at low-altitude in the commercial avocado production area. The ability to distinguish laurel wilt-affected trees from other factors that produce similar external symptoms was also studied. R_modGB digital values of healthy trees and laurel wilt-affected trees, as well as fruit stress and vines covering trees were used to calculate several vegetation indices (VIs), band ratios, and VI combinations. These indices were subjected to analysis of variance (ANOVA) and an M-statistic was performed in order to quantify the separability of those classes. Significant differences in spectral values among laurel wilt affected and healthy trees were observed in all vegetation indices calculated, although the best results were achieved with Excess Red (ExR), (Red–Green) and Combination 1 (COMB1) in all locations. B/G showed a very good potential for separate the other factors with symptoms similar to laurel wilt-affected trees, such as fruit stress and vines covering trees, from laurel wilt-affected trees. These consistent results prove the usefulness of using a modified camera (R_modGB) to discriminate laurel wilt-affected avocado trees from healthy trees, as well as from other factors that cause the same symptoms and suggest performing the classification in further research. According to our results, ExR and B/G should be utilized to develop an algorithm or decision rules to classify aerial images, since they showed the highest capacity to discriminate laurel wilt-affected trees. This methodology may allow the rapid detection of laurel wilt-affected trees using low altitude aerial images and be a valuable tool in mitigating this important threat to Florida avocado production.

Laurel wilt (Lw) is a fatal disease. It is a vascular pathogen and is considered a major threat to the avocado industry in Florida. Many of the symptoms of Lw resemble those that are caused by other diseases or stress factors. In this study, the best wavelengths with which to discriminate plants affected by Lw from stress factors were determined and classified. Visible-near infrared (400–950 nm) spectral data from healthy trees and those with Lw, Phytophthora, or salinity damage were collected using a handheld spectroradiometer. The total number of wavelengths was averaged in two ranges: 10 nm and 40 nm. Three classification methods, stepwise discriminant (STEPDISC) analysis, multilayer perceptron (MLP), and radial basis function (RBF), were applied in the early stage of Lw infestation. The classification results obtained for MLP, with percent accuracy of classification as high as 98% were better than STEPDISC and RBF. The MLP neural network selected certain wavelengths that were crucial for correctly classifying healthy trees from those with stress trees. The results showed that there were sufficient spectral differences between laurel wilt, healthy trees, and trees that have other diseases; therefore, a remote sensing technique could diagnose Lw in the early stage of infestation.

In this study, we combine a fluorescence imaging technique and two machine-learning methods to discriminate Huanglongbing (HLB) disease from zinc-deficiency stress on samples from Florida, USA. Two classification methods, support vector machine (SVM) and artificial neural network (ANN), are used. Our classification results present high accuracy for both classification methods: 92.8% for SVM and 92.2% for ANN. The results from Florida are also compared to results from São Paulo State, Brazil. This comparison indicates that the present technique can be applied to discriminate HLB from zinc deficiency in both states.

Nutrient assessment and management are important to maintain productivity in citrus orchards. In this study, laser-induced breakdown spectroscopy (LIBS) was applied for rapid and real-time detection of citrus anomalies. Laser-induced breakdown spectroscopy spectra were collected from citrus leaves with anomalies such as diseases (Huanglongbing, citrus canker) and nutrient deficiencies (iron, manganese, magnesium, zinc), and compared with those of healthy leaves. Baseline correction, wavelet multivariate denoising, and normalization techniques were applied to the LIBS spectra before analysis. After spectral preprocessing, features were extracted using principal component analysis and classified using two models, quadratic discriminant analysis and support vector machine (SVM). The SVM resulted in a high average classification accuracy of 97.5%, with high average canker classification accuracy (96.5%). LIBS peak analysis indicated that high intensities at 229.7, 247.9, 280.3, 393.5, 397.0, and 769.8 nm were observed of 11 peaks found in all the samples. Future studies using controlled experiments with variable nutrient applications are required for quantification of foliar nutrients by using LIBS-based sensing.

The overall objective of this work was to develop and evaluate computer vision and machine learning technique for classification of Huanglongbing-(HLB)-infected and healthy leaves using fluorescence imaging spectroscopy. The fluorescence images were segmented using normalized graph cut, and texture features were extracted from the segmented images using cooccurrence matrix. The extracted features were used as an input into the classifier, support vector machine (SVM). The classification results were evaluated based on classification accuracies and number of false positives and false negatives. The results indicated that the SVM could classify HLB-infected leaf fluorescence intensities with up to 90% classification accuracy. Though the fluorescence intensities from leaves collected in Brazil and the USA were different, the method shows potential for detecting HLB.