Potato Crop Health Assessment Using Multispectral Image Analysis

Unmanned Aerial Vehicles (UAVs) have become increasingly popular in recent years for agricultural research. High spatial and temporal resolution images obtained with UAVs are ideal for many applications in agriculture. The objective of this study was to evaluate the performance of vegetation indices (VIs) derived from UAV images for quantification of plant nitrogen (N) concentration of spring wheat, a major cereal crop worldwide. This study was conducted at three locations in Idaho, United States. A quadcopter UAV equipped with a red edge multispectral sensor was used to collect images during the 2016 growing season. Flight missions were successfully carried out at Feekes 5 and Feekes 10 growth stages of spring wheat. Plant samples were collected on the same days as UAV image data acquisition and were transferred to lab for N concentration analysis. Different VIs including Normalized Difference Vegetative Index (NDVI), Red Edge Normalized Difference Vegetation Index (NDVIred edge), Enhanced Vegetation Index 2 (EVI2), Red Edge Simple Ratio (SRred edge), Green Chlorophyll Index (CIgreen), Red Edge Chlorophyll Index (CIred edge), Medium Resolution Imaging Spectrometer (MERIS) Terrestrial Chlorophyll Index (MTCI) and Red Edge Triangular Vegetation Index (core only) (RTVIcore) were calculated for each flight event. At Feekes 5 growth stage, red edge and green based VIs showed higher correlation with plant N concentration compare to the red based VIs. At Feekes 10 growth stage, all calculated VIs showed high correlation with plant N concentration. Empirical relationships between VIs and plant N concentration were cross validated using test data sets for each growth stage. At Feekes 5, the plant N concentration estimated based on NDVIred edge showed one to one correlation with measured N concentration. At Feekes 10, the estimated and measured N concentration were highly correlated for all empirical models, but the model based on CIgreen was the only model that had a one to one correlation between estimated and measured plant N concentration. The observed high correlations between VIs derived from UAV and the plant N concentration suggests the significance of VIs deriving from UAVs for within-season N concentration monitoring of agricultural crops such as spring wheat.

Many studies in remote sensing aim to identify varietal differences, predict yield, identify crop nutrition deficiencies, crop area estimates, and other spectral information of crops. Airborne hyperspectral imagery is a potential tool to monitor the nutrient status of citrus trees, since this sensing technology has been shown to efficiently detect nitrogen deficiency in other crops. Therefore, it contributes to save cost, time and labor when used over large areas. The objective of this study was: (i) to determine the spectral behavior of a citrus orchard in high resolution hyperspectral imagery collected with aircraft; (ii) to analyze the correlation between plant parameters and radiometric parameters, especially of the nitrogen nutritional status of citrus orchards. The experiment was conducted in orchards located near Lake Alfred and Auburndale, Polk County, Florida, USA. The variety studied is Valencia (Citrus sinensis). Hyperspectral images consist of 128 visible and near-infrared wavelengths (457.2 – 921.7 nm). Detailed ground truth data were collected at the same time as the imagery for assessing foliar nutrition of citrus groves. The spectral features of the individual trees were identified using pixel-based average spectral reflectance values at several wavelengths from the image. Hyperspectral images and ground truth data were assessed with statistical, GIS and modeling tools, i.e. Vegetation Index (Normalized Difference Vegetation Index, Red Edge Normalized Difference Vegetation Index, Modified Red Edge Simple Ratio, Vogelmann Red Edge Index 1, and others.). The methods Backward Multiple Linear Regression and Principal Component Analysis (PCA) were used to develop prediction models for leaf nitrogen. These results will contribute to the development of rapid detection techniques of nutritional status in citrus groves, in order to reduce the sampling procedures for foliar analysis, which ...

Effects of Band Positioning and Bandwidth on NDVI Measurements of Tropical Savannas

P68. Machine learning algorithms for predicting patient-controlled analgesia use after minimally invasive spine surgery

Globally, subtropical forests are rich in biodiversity. However, the native biodiversity in these forests is threatened by the presence of invasive species such as Chromolaena odorata (L.) King and Robinson, which thrives in forest canopy gaps. Our study explored the utility of WorldView-2 data, an 8-band high resolution (2 m) imagery for mapping the probability of C. odorata occurrence (presence/absence) in canopy gaps of a subtropical forest patch, the Dukuduku forest, South Africa. An integrated modelling approach involving the WorldView-2 vegetation indices and ancillary environmental data was also assessed. The results showed a higher performance of the environmental data only model (deviance or D 2 = 0.52, p < 0.05, n = 77) when compared to modelling with WorldView-2 vegetation indices such as the enhanced vegetation index, simple ratio indices and red edge normalized difference vegetation index (D 2 = 0.30, p < 0.05, n = 77). The integrated model explained the highest presence/absence variance of C. odorata (D 2 = 0.71, i.e. 71 %). This model was used to derive a probability map indicating the occurrence of invasive species in forest gaps. A 2 × 2 error matrix table and the receiver operating characteristic curve derived from an independent validation dataset (n = 38) were used to assess the model accuracy. Approximately 87 % of canopy gaps containing C. odorata were correctly predicted at probability threshold of 0.3. The derived probability map of C. odorata occurrence could assist management in prioritizing target areas for eradication of the species.

As pressure on available water resources increases, the need to exploit technology in order to produce more food with less water becomes crucial. The technological hardware requisite for precise water delivery methods, such as variable rate irrigation, is commercially available. Despite that, techniques to formulate a timely, accurate prescription for those systems are inadequate. Spectral vegetation indices are often used to gauge crop vigor and related parameters (e.g., leaf nitrogen [N] content). However, research heretofore rarely addresses the influence of soil moisture on the indices. The objectives of this study were to determine (1) if vegetation indices derived from multispectral satellite imagery could assist in quantifying soil moisture variability in irrigated maize (<i>Zea mays</i> L.) production, and (2) the period of time that a single image is representative of soil moisture. A variable rate irrigation pivot was used to form six water treatment zones. Each was equipped with tensiometers installed in the center of the plots at 20, 45, and 75 cm depths to individually monitor conditions in the water treatment zones. Water was applied for each treatment as a percentage of the estimated evapotranspiration (ET) requirement: 40%, 60%, 80%, 100%, 120%, and 140%. Data collected from tensiometers was paired with the image pixels corresponding to its ground location. Statistical analysis was performed separately to assess whether vegetation indices are representative of soil moisture at several crop growth stages. Findings from this study indicate that Red Edge Normalized Difference Vegetation Index could quantify soil moisture tension variability at V6 (six leaf; <i>r</i>² = 0.850, <i>p</i> = 0.009) and V9 (nine leaf; <i>r</i>² = 0.913, <i>p</i> = 0.003) crop growth stages. Results suggest that satellite-derived vegetation indices may be useful for creating time-sensitive characterizations of soil moisture variability. Further study is necessary to investigate additional crop growth stages, more crops, and other sources of multispectral imagery.

Leaf rust (Puccinia triticina Eriks) is a wheat disease causing substantial yield losses in wheat production globally. The identification of genetic resources with permanently effective resistance genes and the generation of mutant lines showing increased levels of resistance allow the efficient incorporation of these target genes into germplasm pools by marker-assisted breeding. In this study, new mutant (M3 generation) lines generated from the rust-resistant variety Kazakhstanskaya-19 were developed using gamma-induced mutagenesis through 300-, 350-, and 400-Gy doses. In field trials after leaf rust inoculation, 75 mutant lines showed adult plant resistance. These lines were evaluated for resistance at the seedling stage via microscopy in greenhouse experiments. Most of these lines (89.33%) were characterized as resistant at both developmental stages. Hyperspectral imaging analysis indicated that infected leaves of wheat genotypes showed increased relative reflectance in visible and near-infrared light compared to the non-infected genotypes, with peak means at 462 and 644 nm, and 1936 and 2392 nm, respectively. Five spectral indexes, including red edge normalized difference vegetation index (RNDVI), structure-insensitive pigment index (SIPI), ratio vegetation index (RVSI), water index (WI), and normalized difference water index (NDWI), demonstrated significant potential for determining disease severity at the seedling stage. The most significant differences in reflectance between susceptible and resistant mutant lines appeared at 694.57 and 987.51 nm. The mutant lines developed were also used for the development and validation of KASP markers for leaf rust resistance genes Lr1, Lr2a, Lr3, Lr9, Lr10, and Lr17. The mutant lines had high frequencies of "a" resistance alleles (0.88) in all six Lr genes, which were significantly associated with seedling resistance and suggest the potential of favorable haplotype introgression through functional markers. Nine mutant lines characterized by the presence of "b" alleles in Lr9 and Lr10-except for one line with allele "a" in Lr9 and three mutant lines with allele "a" in Lr10-showed the progressive development of fungal haustorial mother cells 72 h after inoculation. One line from 300-Gy-dosed mutant germplasm with "b" alleles in Lr1, Lr2a, Lr10, and Lr17 and "a" alleles in Lr3 and Lr9 was characterized as resistant based on the low number of haustorial mother cells, suggesting the contribution of the "a" alleles of Lr3 and Lr9.

Vegetation indices obtained from radiometric measurements have been used to estimate the stress response of plants grown in contaminated sites. The phytotoxicity of Pb and Zn in Festuca rubra L. and Vulpia myuros L. plants grown under hydroponic conditions was evaluated using vegetation indices obtained from radiometric measurements. The plants were supplied with 3mM Zn (+Zn), 500μM Pb (+Pb) and 500μM Pb with EDTA (+PbEDTA) for 3months. Significantly higher Zn concentrations in F. rubra shoots compared with V. myuros shoots were detected for Zn and Pb treatments. EDTA increased Pb transport to the shoots for both grasses, while Pb-treated plants retained Pb primarily in the roots. All vegetation indices tested showed the highest differences in F. rubra under +PbEDTA treatment and minor effects under +Zn, whereas the major variations for V. myuros corresponded to +Zn treatment, followed by +PbEDTA. Red edge normalized difference vegetation index, yellowness index and anthocyanin concentration index were the most sensitive indices to report Zn and Pb phytotoxicity in these grasses. According to the results obtained, both metal concentrations and radiometric indices suggested that Pb is more phytotoxic to F. rubra, which tolerates high Zn levels, whereas V. myuros was strongly affected by high Zn levels and markedly tolerant to Pb, even when applied in a mobile form (PbEDTA). Both species could be used in the phytostabilization of Zn- and Pb-contaminated soils. The abilities of F. rubra to accumulate Zn and V. myuros to accumulate Pb in the roots would facilitate a more efficient phytoremediation strategy when used in combination.

In recent years, there have been rapid improvements in both remote sensing methods and satellite image availability that have the potential to massively improve burn severity assessments of the Alaskan boreal forest. In this study, we utilized recent pre- and post-fire Sentinel-2 satellite imagery of the 2019 Nugget Creek and Shovel Creek burn scars located in Interior Alaska to both assess burn severity across the burn scars and test the effectiveness of several remote sensing methods for generating accurate map products: Normalized Difference Vegetation Index (NDVI), Normalized Burn Ratio (NBR), and Random Forest (RF) and Support Vector Machine (SVM) supervised classification. We used 52 Composite Burn Index (CBI) plots from the Shovel Creek burn scar and 28 from the Nugget Creek burn scar for training classifiers and product validation. For the Shovel Creek burn scar, the RF and SVM machine learning (ML) classification methods outperformed the traditional spectral indices that use linear regression to separate burn severity classes (RF and SVM accuracy, 83.33%, versus NBR accuracy, 73.08%). However, for the Nugget Creek burn scar, the NDVI product (accuracy: 96%) outperformed the other indices and ML classifiers. In this study, we demonstrated that when sufficient ground truth data is available, the ML classifiers can be very effective for reliable mapping of burn severity in the Alaskan boreal forest. Since the performance of ML classifiers are dependent on the quantity of ground truth data, when sufficient ground truth data is available, the ML classification methods would be better at assessing burn severity, whereas with limited ground truth data the traditional spectral indices would be better suited. We also looked at the relationship between burn severity, fuel type, and topography (aspect and slope) and found that the relationship is site-dependent.

Este trabalho foi realizado para determinar, entre seis índices de vegetação baseados em dados de refletância espectral, aquele que melhor discrimina doses de nitrogênio e que possui maior correlação com leituras de clorofila e massa seca do capim-tanzânia (Panicum maximum Jacq.). Os índices testados foram o NDVI (índice de vegetação por diferença normalizada, calculado utilizando a banda do vermelho e a banda do infravermelho próximo), VARI (índice resistente à atmosfera na região do visível, calculado utilizando a banda de transição do vermelho ao infravermelho e a banda do verde) e WDRVI (índice de vegetação de amplo alcance, calculado utilizando três coeficientes de ponderação, 0,05; 0,1 e 0,2). Foram avaliadas quatro doses de nitrogênio (0, 80, 160 e 320 kg/ha) em delineamento de blocos ao acaso com subamostras, com três repetições e três subamostras por bloco. Os índices VARI utilizando a banda de transição do vermelho ao infravermelho e o WDRVI utilizando os coeficientes 0,05 e 0,1 foram os melhores para discriminar as doses de nitrogênio aplicadas em todos os períodos de avaliação estudados. O índice que apresentou maior correlação com as leituras de clorofila e massa seca foi o VARI com banda de transição do vermelho ao infravermelho.

Parkinson’s disease (PD) is a nervous system disorder first described as a neurological condition in 1817. It is one of the more prevalent diseases in the elderly, and Alzheimer’s is the second most common neurodegenerative illness. It impacts the patient’s movement. Symptoms start gradually with tremors, stiffness in movement, and speech and voice disorders. Researches proved that 89% of patients with Parkinson’s has speech disorder including uncertain articulation, hoarse and breathy voice and monotone pitch. The cause behind this voice change is the reduction of dopamine due to damage of neurons in the substantia nigra responsible for dopamine production. In this work, Parkinson’s disease is classified with the help of human voice signals. Six different machine learning (ML) algorithms are used in the classification: Stochastic Gradient Descent (SGD) Classifier, Extreme Gradient Boosting (XGB) Classifier, Logistic Regression Classifier, Random Forest Classifier, K-Nearest Neighbour (KNN) Classifier, and Decision Tree (DT) Classifier. This research aims to classify Parkinson’s disease using human voice signals and extract essential features to reduce the complexity of the dataset. Then, human voice signals are analyzed to check the voice intensity and spectrum for PD patients. Then, machine learning classifiers are applied to classify the PD patients based on the extracted features. The results show that SGD-Classifier has 91% accuracy, XGB-Classifier has 95% accuracy, Logistic Regression has 91% accuracy, Random Forest shows 97% accuracy, KNN shows 95% accuracy, and Decision Tree has 95% accuracy. Hence, Random Forest has the highest accuracy. The disease can be studied more by looking for more characteristics of PD patients to enhance its proper use in the medical field.

The design of this system shows the efficient prediction of heart diseases by the implementation of machine learning multiple classifiers algorithm. The UCI Heart Disease Dataset is used, while the data preprocessing involves filling missing values, encoding of categorical values, scaling of features, and selection of important features. Three machine learning classifiers, namely Random Forest Classifier, Gradient Boosting Classifier, and Logistic Regression are implemented. The implementation involves training the machine learning models using the RandomizedSearchCV technique and cross-validation technique. The evaluation of the model's performance is done using measures such as ACCURACY, PRECISION, RECALL, and F1-SCORE. A fair comparison is conducted to select the best performing model automatically using accuracy. Results show that the Random Forest model performed the best with an accuracy of 81.67%, compared to the rest of the models. Evaluation of feature importance shows key clinical features that significantly contribute to the accuracy. Keywords— Heart Disease Prediction, Machine Learning, Random Forest, Gradient Boosting, Logistic Regression, Feature Selection, Classification, Healthcare Analytics, UCI Dataset, Predictive Modeling

The early detection and rapid identification of life-threatening diseases play crucial roles in saving lives. However, disparities in healthcare infrastructure and the unequal distribution of medical resources often hinder timely diagnosis, resulting in adverse outcomes. This study introduces a Multi-Disease Prediction System, a user-friendly web application developed with Streamlit, capable of predicting kidney disease, breast cancer, and heart disease. Unlike traditional AI models that focus on predicting a single disease, our objective is to develop a unified platform using Streamlit and Machine learning to predict multiple diseases simultaneously. The system harnesses advanced machine learning techniques, utilizing algorithms such as the Random Forest, Gradient Boosting Classifier, XGBoost, K-Nearest Neighbor, and Decision Tree classifier. For each disease, the optimal algorithm was selected based on its performance, yielding predictive accuracies of 99% for kidney disease (Random Forest Classifier), 98% for breast cancer (XGBoost), and 95% for heart disease (Random Forest Classifier). Users can input health indicators and symptoms and receive personalized risk insights. By offering an intuitive and interactive platform for data-driven screening, this application seeks to enhance early detection and encourage proactive health management in diverse populations. These models are designed to continuously improve through advancements in machine learning, ensuring reliable and impactful health predictions for broader accessibility and preventive care. Additionally, future improvements, such as integrating more diseases into the system, would further enhance its capabilities and expand its impact on early disease detection.

The quality of Earth observation (EO) based vegetation monitoring has improved during recent years, which can be attributed to the enhanced sensor design of new satellites such as MODIS (Moderate Resolution Imaging Spectroradiometer) on Terra and Aqua. It is however expected that sun-sensor geometry variations will have a more visible impact on the Normalized Difference Vegetation Index (NDVI) from MODIS compared to earlier data sources, since noise related to atmosphere and sensor calibration is substantially reduced in the MODIS data stream. For this reason, the effect of varying MODIS viewing geometry on red, near-infrared (NIR) and NDVI needs to be quantified. Data from the geostationary MSG (Meteosat Second Generation) SEVIRI (Spinning Enhanced Visible and Infrared Imager) sensor is well suited for this purpose due to the fixed position of the sensor, the spectral resolution, including a red and NIR band, and the high temporal resolution (15 min) of data, enabling MSG data to be used as a reference for estimating MODIS surface reflectance and NDVI variations caused by varying sun-sensor geometry. The study was performed on data covering West Africa for periods of lowest possible cloud cover for three consecutive years (2004–2006). An analysis covering the entire range of NDVI revealed day-to-day variations in observed MODIS NDVI of 50–60% for medium dense vegetation (NDVI ≈ 0.5) caused by variations in MODIS view zenith angles (VZAs) between nadir and the high forward-scatter view direction. Statistical analysis on red, NIR and NDVI from MODIS and MSG SEVIRI for three transects (characterized by different vegetation densities) showed that both MODIS red and NIR reflectances are highly dependant on MODIS VZA and relative azimuth angle (RAA), due to the anisotropic behaviour of red and NIR reflectances. The anisotropic reflectance in the red and NIR band was to some degree minimized by the ratioing properties of NDVI. The minimization by the NDVI normalization is very dependent on the vegetation density however, since the degree of anisotropy in red and NIR reflectances depends on the amount of vegetation present. MODIS VZA and RAA effects on NDVI were highest for medium dense vegetation (NDVI ≈ 0.5–0.6). The VZA and RAA effects were less for sparsely vegetated areas (NDVI ≈ 0.3–0.35) and the smallest effect on NDVI was found for dense vegetation (NDVI ≈ 0.7). These results have implications for the end users' interpretation of NDVI, and challenge the expediency of the MODIS NDVI compositing technique, which should be refined to distinguish between forward- and backward-scatter viewing direction by taking RAA into account.

Cropland mapping using remote sensing data is the basis for effective crop monitoring, crop rotation control, and the detection of irrational land use. Classification using Normalized Difference Vegetation Index (NDVI) time series from multi-year data requires additional time costs, especially when sentinel data are sparse. Approximation by nonlinear functions was proposed to solve this problem. Time series of weekly NDVI composites were plotted using multispectral Sentinel-2 (Level-2A) images at a resolution of 10 m for sites in Khabarovsk District from April to October in the years 2021 and 2022. Missing values due to the lack of suitable images for analysis were recovered using cubic polynomial, Fourier series, and double sinusoidal function approximation. The classes that were considered included crops, namely, soybean, buckwheat, oat, and perennial grasses, and fallow. The mean absolute percentage error (MAPE) of each class fitting was calculated. It was found that Fourier series fitting showed the highest accuracy, with a mean error of 8.2%. Different classifiers, such as the support vector machine (SVM), random forest (RF), and gradient boosting (GB), were comparatively evaluated. The overall accuracy (OA) for the site pixels during the cross-validation (Fourier series restored) was 67.3%, 87.2%, and 85.9% for the SVM, RF, and GB classifiers, respectively. Thus, it was established that the best result in terms of combined accuracy, performance, and limitations in cropland mapping was achieved by composite construction using Fourier series and machine learning using GB. Similar results should be expected in regions with similar cropland structures and crop phenological cycles, including other regions of the Far East.