Quantitative Assessment of Apple Mosaic Disease Severity Based on Hyperspectral Images and Chlorophyll Content

Remote Sensing, Journal Year: 2023, Volume and Issue: 15(8), P. 2202 - 2202

Published: April 21, 2023

The infection of Apple mosaic virus (ApMV) can severely damage the cellular structure apple leaves, leading to a decrease in leaf chlorophyll content (LCC) and reduced fruit yield. In this study, we propose novel method that utilizes hyperspectral imaging (HSI) technology non-destructively monitor ApMV-infected leaves predict LCC as quantitative indicator disease severity. data were collected from 360 optimal wavelengths selected using competitive adaptive reweighted sampling algorithms. A high-precision inversion model was constructed based on Boosting Stacking strategies, with validation set Rv2 0.9644, outperforming traditional ensemble learning models. used invert distribution image calculate average coefficient variation (CV) for each leaf. Our findings indicate CV highly correlated severity, their combination sensitive enabled accurate identification severity (validation overall accuracy = 98.89%). approach considers role plant chemical composition provides comprehensive evaluation at scale. Overall, our study presents an effective way evaluate health status offering quantifiable index aid prevention control.

Language: Английский

---

Estimating soil salinity in mulched cotton fields using UAV-based hyperspectral remote sensing and a Seagull Optimization Algorithm-Enhanced Random Forest Model

Computers and Electronics in Agriculture, Journal Year: 2024, Volume and Issue: 221, P. 109017 - 109017

Published: May 7, 2024

Language: Английский

---

Quantifying corn LAI using machine learning and UAV multispectral imaging

Precision Agriculture, Journal Year: 2024, Volume and Issue: 25(4), P. 1777 - 1799

Published: March 28, 2024

Language: Английский

---

Inversion of Leaf Chlorophyll Content in Different Growth Periods of Maize Based on Multi-Source Data from “Sky–Space–Ground”

Remote Sensing, Journal Year: 2025, Volume and Issue: 17(4), P. 572 - 572

Published: Feb. 8, 2025

Leaf chlorophyll content (LCC) is a key indicator of crop growth condition. Real-time, non-destructive, rapid, and accurate LCC monitoring paramount importance for precision agriculture management. This study proposes an improved method based on multi-source data, combining the Sentinel-2A spectral response function (SRF) computer algorithms, to overcome limitations traditional methods. First, equivalent remote sensing reflectance was simulated by UAV hyperspectral images with ground experimental data. Then, using grey relational analysis (GRA) maximum information coefficient (MIC) algorithm, we explored complex relationship between vegetation indices (VIs) LCC, further selected feature variables. Meanwhile, utilized three (DSI, NDSI, RSI) identify sensitive band combinations analyzed original bands LCC. On this basis, nonlinear machine learning models (XGBoost, RFR, SVR) one multiple linear regression model (PLSR) construct inversion model, chose optimal generate spatial distribution maps maize at regional scale. The results indicate that there significant correlation VIs XGBoost, SVR outperforming PLSR model. Among them, XGBoost_MIC achieved best during tasseling stage (VT) growth. In R2 = 0.962 RMSE 5.590 mg/m2 in training set, 0.582 6.019 test set. For Sentinel-2A-simulated set had 0.923 8.097 mg/m2, while showed 0.837 3.250 which indicates improvement accuracy. scale, also yielded good (train 0.76, 0.88, 18.83 mg/m2). conclusion, proposed not only significantly improves accuracy methods but also, its outstanding versatility, can achieve precise different regions various types, demonstrating broad application prospects practical value agriculture.

Language: Английский

---

Hyperspectral Estimation of Chlorophyll Content in Apple Tree Leaf Based on Feature Band Selection and the CatBoost Model

Agronomy, Journal Year: 2023, Volume and Issue: 13(8), P. 2075 - 2075

Published: Aug. 7, 2023

Leaf chlorophyll content (LCC) is a crucial indicator of nutrition in apple trees and can be applied to assess their growth status. Hyperspectral data provide an important means for detecting the LCC trees. In this study, hyperspectral measured were obtained. The original spectrum (OR) was pretreated using some spectral transformations. Feature bands selected based on competitive adaptive reweighted sampling (CARS) algorithm, random frog (RF) elastic net (EN) EN-RF EN-CARS algorithms. Partial least squares regression (PLSR), forest (RFR), CatBoost algorithm used before after grid search parameter optimization estimate LCC. results revealed following: (1) second derivative (SD) transformation had highest correlation with (–0.929); moreover, SD-based model produced accuracy, making SD effective pretreatment method tree estimation. (2) Compared single band selection better dimension reduction effect, modeling accuracy generally higher. (3) best estimation validation set SD-EN-CARS-CatBoost determination coefficient (R2), root mean square error (RMSE), relative prediction deviation (RPD) reaching 0.923, 2.472, 3.64, respectively. As such, optimized model, its high reliability, monitor trees, support intelligent management orchards, facilitate economic development fruit industry.

Language: Английский

---

Enhancing phenotyping efficiency in faba bean breeding: integrating UAV imaging and machine learning

Precision Agriculture, Journal Year: 2024, Volume and Issue: 25(3), P. 1502 - 1528

Published: March 6, 2024

Abstract Unmanned aerial vehicles (UAVs) equipped with high-resolution imaging sensors have shown great potential for plant phenotyping in agricultural research. This study aimed to explore the of UAV-derived red–green–blue (RGB) and multispectral data estimating classical measures such as height predicting yield chlorophyll content (indicated by SPAD values) a field trial 38 faba bean ( Vicia L.) cultivars grown at four replicates south-eastern Norway. To predict values, Support Vector Regression (SVR) Random Forest (RF) models were utilized. Two feature selection methods, namely Pearson correlation coefficient (PCC) sequential forward (SFS), applied identify most relevant features prediction. The incorporated various combinations bands, indices, UAV-based values different development stages. between manual measurements revealed strong agreement (R 2 ) 0.97. best prediction value was achieved BBCH 50 (flower bud present) an R 0.38 RMSE 1.14. For prediction, 60 (first flower open) identified optimal stage, using spectral indices yielding 0.83 0.53 tons/ha. stage presents opportunity implement targeted management practices enhance yield. integration UAVs RGB cameras, along machine learning algorithms, proved be accurate approach agronomically important traits bean. methodology offers practical solution rapid efficient high-throughput breeding programs.

Language: Английский

---

Grape leaf moisture prediction from UAVs using multimodal data fusion and machine learning

Precision Agriculture, Journal Year: 2024, Volume and Issue: 25(3), P. 1609 - 1635

Published: March 25, 2024

Language: Английский

---

Hyperspectral Estimation of SPAD Value of Cotton Leaves under Verticillium Wilt Stress Based on GWO–ELM

Agriculture, Journal Year: 2023, Volume and Issue: 13(9), P. 1779 - 1779

Published: Sept. 7, 2023

Rapid and non-destructive estimation of the chlorophyll content in cotton leaves is great significance for real-time monitoring growth under verticillium wilt (VW) stress. The spectral reflectance healthy VW was determined using hyperspectral technology, original spectra were processed Savitzky–Golay (SG) smoothing, on its basis through mean centering, standard normal variate (SG-SNV), multiplicative scatter correction (SG-MSC), reciprocal second-order differentiation, logarithmic differentiation ([lg(SG)]″) preprocessing operations. characteristic bands selected based correlation coefficient, vegetation index, successive projection algorithm (SPA), competitive adaptive reweighted sampling (CARS). single-factor model, back propagation neural network particle swarm optimization algorithm, extreme learning machine (ELM) a grey wolf optimizer (GWO) constructed to compare explore ability each model estimate soil plant analysis development (SPAD) value results showed that pretreatment could improve between SPAD values. SG-MSC SG-SNV better changes five pretreatments, maximum coefficients higher than 0.74. Compared with SPA, accuracy CARS-extracted higher, multi-factor pretreatment. For leaves, [lg(SG)]″–CARS–GWO–ELM optimal modeling validation set R2 0.956 0.887, respectively. SG-MSC–CARS–GWO–ELM 0.832 0.824, Therefore, GWO–ELM different pretreatments combined extraction methods can be used leaf values stress dynamically monitor provide theoretical reference precision agriculture.

Language: Английский

---

Ground-Based Hyperspectral Estimation of Maize Leaf Chlorophyll Content Considering Phenological Characteristics

Remote Sensing, Journal Year: 2024, Volume and Issue: 16(12), P. 2133 - 2133

Published: June 13, 2024

Accurately measuring leaf chlorophyll content (LCC) is crucial for monitoring maize growth. This study aims to rapidly and non-destructively estimate the LCC during four critical growth stages investigate ability of phenological parameters (PPs) LCC. First, spectra were obtained by spectral denoising followed transformation. Next, sensitive bands (Rλ), indices (SIs), PPs extracted from all at each stage. Then, univariate models constructed determine their potential independent estimation. The multivariate regression (LCC-MR) built based on SIs, SIs + Rλ, Rλ after feature variable selection. results indicate that our machine-learning-based LCC-MR demonstrated high overall accuracy. Notably, 83.33% 58.33% these showed improved accuracy when successively introduced SIs. Additionally, model accuracies milk-ripe tasseling outperformed those flare–opening jointing under identical conditions. optimal was created using XGBoost, incorporating SI, PP variables R3 These findings will provide guidance support management.

Language: Английский

---

Hyperspectral estimation of chlorophyll density in winter wheat using fractional-order derivative combined with machine learning

Frontiers in Plant Science, Journal Year: 2025, Volume and Issue: 15

Published: Jan. 14, 2025

Chlorophyll density (ChD) can reflect the photosynthetic capacity of winter wheat population, therefore achieving real-time non-destructive monitoring ChD in is great significance for evaluating growth status wheat. Derivative preprocessing has a wide range applications hyperspectral chlorophyll. In order to research role fractional-order derivative (FOD) model ChD, this study based on an irrigation experiment obtain and canopy reflectance. The original spectral reflectance curves were preprocessed using 3 FOD methods: Grünwald-Letnikov (GL), Riemann-Liouville (RL), Caputo. Hyperspectral models constructed 8 machine learning algorithms, including partial least squares regression, support vector multi-layer perceptron random forest extra-trees regression (ETsR), decision tree K-nearest neighbors gaussian process full spectrum band selected by competitive adaptive reweighted sampling (CARS). main results as follows: For types FOD, GL-FOD was suitable analyzing change curve towards integer-order curve. RL-FOD constructing ChD. Caputo-FOD not due its insensitivity changes order. calculation methods could all improve correlation between Log(ChD) varying degrees, but only GL method RL observe with changes, shorter wavelength, smaller order, higher correlation. bands screened CARS distributed throughout entire range, there relatively concentrated distribution visible light region. Among models, used screen 0.3-order spectrum, ETsR reached best accuracy stability. Its R 2c , RMSE c 2v v RPD 1.0000, 0.0000, 0.8667, 0.1732, 2.6660, respectively. conclusion, data set corresponding set, combined methods, 1 screening method, modeling showed that RL-FOD, band, highest accuracy, estimation be realized. provide some reference rapid nondestructive

Language: Английский

---