Unmanned Aerial Vehicle Multispectral Images Research Articles

Inversion of winter wheat leaf area index from UAV multispectral images: classical vs. deep learning approaches.

Precise and timely leaf area index (LAI) estimation for winter wheat is crucial for precision agriculture. The emergence of high-resolution unmanned aerial vehicle (UAV) data and machine learning techniques offers a revolutionary approach for fine-scale estimation of wheat LAI at the low cost. While machine learning has proven valuable for LAI estimation, there are still model limitations and variations that impede accurate and efficient LAI inversion. This study explores the potential of classical machine learning models and deep learning model for estimating winter wheat LAI using multispectral images acquired by drones. Initially, the texture features and vegetation indices served as inputs for the partial least squares regression (PLSR) model and random forest (RF) model. Then, the ground-measured LAI data were combined to invert winter wheat LAI. In contrast, this study also employed a convolutional neural network (CNN) model that solely utilizes the cropped original image for LAI estimation. The results show that vegetation indices outperform the texture features in terms of correlation analysis with LAI and estimation accuracy. However, the highest accuracy is achieved by combining both vegetation indices and texture features to invert LAI in both conventional machine learning methods. Among the three models, the CNN approach yielded the highest LAI estimation accuracy (R 2 = 0.83), followed by the RF model (R 2 = 0.82), with the PLSR model exhibited the lowest accuracy (R 2 = 0.78). The spatial distribution and values of the estimated results for the RF and CNN models are similar, whereas the PLSR model differs significantly from the first two models. This study achieves rapid and accurate winter wheat LAI estimation using classical machine learning and deep learning methods. The findings can serve as a reference for real-time wheat growth monitoring and field management practices.

Remote estimation of rapeseed phenotypic traits under different crop conditions based on unmanned aerial vehicle multispectral images

Remote estimation of rapeseed phenotypic traits under different crop conditions based on unmanned aerial vehicle multispectral images

Estimation of the Bio-Parameters of Winter Wheat by Combining Feature Selection with Machine Learning Using Multi-Temporal Unmanned Aerial Vehicle Multispectral Images

Accurate and timely monitoring of biochemical and biophysical traits associated with crop growth is essential for indicating crop growth status and yield prediction for precise field management. This study evaluated the application of three combinations of feature selection and machine learning regression techniques based on unmanned aerial vehicle (UAV) multispectral images for estimating the bio-parameters, including leaf area index (LAI), leaf chlorophyll content (LCC), and canopy chlorophyll content (CCC), at key growth stages of winter wheat. The performance of Support Vector Regression (SVR) in combination with Sequential Forward Selection (SFS) for the bio-parameters estimation was compared with that of Least Absolute Shrinkage and Selection Operator (LASSO) regression and Random Forest (RF) regression with internal feature selectors. A consumer-grade multispectral UAV was used to conduct four flight campaigns over a split-plot experimental field with various nitrogen fertilizer treatments during a growing season of winter wheat. Eighteen spectral variables were used as the input candidates for analyses against the three bio-parameters at four growth stages. Compared to LASSO and RF internal feature selectors, the SFS algorithm selects the least input variables for each crop bio-parameter model, which can reduce data redundancy while improving model efficiency. The results of the SFS-SVR method show better accuracy and robustness in predicting winter wheat bio-parameter traits during the four growth stages. The regression model developed based on SFS-SVR for LAI, LCC, and CCC, had the best predictive accuracy in terms of coefficients of determination (R2), root mean square error (RMSE) and relative predictive deviation (RPD) of 0.967, 0.225 and 4.905 at the early filling stage, 0.912, 2.711 μg/cm2 and 2.872 at the heading stage, and 0.968, 0.147 g/m2 and 5.279 at the booting stage, respectively. Furthermore, the spatial distributions in the retrieved winter wheat bio-parameter maps accurately depicted the application of the fertilization treatments across the experimental field, and further statistical analysis revealed the variations in the bio-parameters and yield under different nitrogen fertilization treatments. This study provides a reference for monitoring and estimating winter wheat bio-parameters based on UAV multispectral imagery during specific crop phenology periods.

Wheat Yield Estimation Based on Unmanned Aerial Vehicle Multispectral Images and Texture Feature Indices

To obtain timely, accurate, and reliable information on wheat yield dynamics. The UAV DJI Wizard 4-multispectral version was utilized to acquire multispectral images of winter wheat during the tasseling, grouting, and ripening periods, and to manually acquire ground yield data. Sixteen vegetation indices were screened by correlation analysis, and eight textural features were extracted from five single bands in three fertility periods. Subsequently, models for estimating winter wheat yield were developed utilizing multiple linear regression (MLR), partial least squares (PLS), BP neural network (BPNN), and random forest regression (RF), respectively. (1) The results indicated a consistent correlation between the two variable types and yield across various fertility periods. This correlation consistently followed a sequence: heading period > filling period > mature stage. (2) The model’s accuracy improves significantly when incorporating both texture features and vegetation indices for estimation, surpassing the accuracy achieved through the estimation of a single variable type. (3) Among the various models considered, the partial least squares (PLS) model integrating texture features and vegetation indices exhibited the highest accuracy in estimating winter wheat yield. It achieved a coefficient of determination (R2) of 0.852, a root mean square error (RMSE) of 74.469 kg·hm−2, and a normalized root mean square error (NRMSE) of 7.41%. This study validates the significance of utilizing image texture features along with vegetation indices to enhance the accuracy of models estimating winter wheat yield. It demonstrates that UAV multispectral images can effectively establish a yield estimation model. Combining vegetation indices and texture features results in a more accurate and predictive model compared to using a single index.

Estimating aboveground biomass of grassland in central Asia mountainous areas using unmanned aerial vehicle vegetation indices and image textures – A case study of typical grassland in Tajikistan

Climate change and human activities have severely impacted Central Asia's mountainous grasslands' health status, making monitoring aboveground biomass (AGB) crucial for grassland protection. However, a high-resolution and low-cost solution for AGB monitoring is lacking for Central Asia's grassland. This research proposes an unmanned aerial vehicle (UAV) based AGB monitoring framework using consumer-level cameras. Texture features from UAV RGB images and vegetation indices from UAV multispectral images are used to predict AGB. As one of the typical mountainous countries in Central Asia, Tajikistan is chosen as the study area. Four different grassland types were chosen to investigate the performance of applying UAV for AGB monitoring. Firstly, correlation analysis was performed to identify important features for AGB estimation. Subsequently, the ground-measured AGB and UAV image features were utilized in multiple linear regression (MLR) and generalized additive model (GAM) to develop an AGB prediction model. The results show that the angular second moment (ASM) of the green band and green normalized difference vegetation index (GNDVI) are the two most important features for AGB estimation in the study area. Furthermore, the GAM-based model demonstrated higher accuracy (R2 = 0.87, RMSE = 127.53 g/m2, rRMSE = 0.17) compared to MLR (R2 = 0.74, RMSE = 181.68 g/m2, rRMSE = 0.24), highlighting the nonlinear relationship between AGB and UAV image features. This research uses multispectral and RGB images to achieve a high accuracy and efficient AGB monitoring framework for Central Asia grasslands and provides a reference to validate satellite images for providing long-term and large-scale monitoring of grassland ecosystems and reasonable utilization of grassland resources.

Combining spectral and texture feature of UAV image with plant height to improve LAI estimation of winter wheat at jointing stage.

Leaf area index (LAI) is a critical physiological and biochemical parameter that profoundly affects vegetation growth. Accurately estimating the LAI for winter wheat during jointing stage is particularly important for monitoring wheat growth status and optimizing variable fertilization decisions. Recently, unmanned aerial vehicle (UAV) data and machine/depth learning methods are widely used in crop growth parameter estimation. In traditional methods, vegetation indices (VI) and texture are usually to estimate LAI. Plant Height (PH) unlike them, contains information about the vertical structure of plants, which should be consider. Taking Xixingdian Township, Cangzhou City, Hebei Province, China as the research area in this paper, and four machine learning algorithms, namely, support vector machine(SVM), back propagation neural network (BPNN), random forest (RF), extreme gradient boosting (XGBoost), and two deep learning algorithms, namely, convolutional neural network (CNN) and long short-term memory neural network (LSTM), were applied to estimate LAI of winter wheat at jointing stage by integrating the spectral and texture features as well as the plant height information from UAV multispectral images. Initially, Digital Surface Model (DSM) and Digital Orthophoto Map (DOM) were generated. Subsequently, the PH, VI and texture features were extracted, and the texture indices (TI) was further constructed. The measured LAI on the ground were collected for the same period and calculated its Pearson correlation coefficient with PH, VI and TI to pick the feature variables with high correlation. The VI, TI, PH and fusion were considered as the independent features, and the sample set partitioning based on joint x-y distance (SPXY) method was used to divide the calibration set and validation set of samples. The ability of different inputs and algorithms to estimate winter wheat LAI were evaluated. The results showed that (1) The addition of PH as a feature variable significantly improved the accuracy of the LAI estimation, indicating that wheat plant height played a vital role as a supplementary parameter for LAI inversion modeling based on traditional indices; (2) The combination of texture features, including normalized difference texture indices (NDTI), difference texture indices (DTI), and ratio texture indices (RTI), substantially improved the correlation between texture features and LAI; Furthermore, multi-feature combinations of VI, TI, and PH exhibited superior capability in estimating LAI for winter wheat; (3) Six regression algorithms have achieved high accuracy in estimating LAI, among which the XGBoost algorithm estimated winter wheat LAI with the highest overall accuracy and best results, achieving the highest R2 (R2=0.88), the lowest RMSE (RMSE=0.69), and an RPD greater than 2 (RPD=2.54). This study provided compelling evidence that utilizing XGBoost and integrating spectral, texture, and plant height information extracted from UAV data can accurately monitor LAI during the jointing stage of winter wheat. The research results will provide a new perspective for accurate monitoring of crop parameters through remote sensing.

Combining Texture, Color, and Vegetation Index from Unmanned Aerial Vehicle Multispectral Images to Estimate Winter Wheat Leaf Area Index during the Vegetative Growth Stage

Leaf Area Index (LAI) is a fundamental indicator of plant growth status in agronomy and environmental research. With the rapid development of drone technology, the estimation of crop LAI based on drone imagery and vegetation indices is becoming increasingly popular. However, there is still a lack of detailed research on the feasibility of using image texture to estimate LAI and the impact of combining texture indices with vegetation indices on LAI estimation accuracy. In this study, two key growth stages of winter wheat (i.e., the stages of green-up and jointing) were selected, and LAI was calculated using digital hemispherical photography. The feasibility of predicting winter wheat LAI was explored under three conditions: vegetation index, texture index, and a combination of vegetation index and texture index, at flight heights of 20 m and 40 m. Two feature selection methods (Lasso and recursive feature elimination) were combined with four machine learning regression models (multiple linear regression, random forest, support vector machine, and backpropagation neural network). The results showed that during the vegetative growth stage of winter wheat, the model combining texture information with vegetation indices performed better than the models using vegetation indices alone or texture information alone. Among them, the best prediction result based on vegetation index was RFECV-MLR at a flight height of 40 m (R2 = 0.8943, RMSE = 0.4139, RRMSE = 0.1304, RPD = 3.0763); the best prediction result based on texture index was RFECV-RF at a flight height of 40 m (R2 = 0.8894, RMSE = 0.4236, RRMSE = 0.1335, RPD = 3.0063); and the best prediction result combining texture and index was RFECV-RF at a flight height of 40 m (R2 = 0.9210, RMSE = 0.3579, RRMSE = 0.1128, RPD = 3.5575). The results of this study demonstrate that combining vegetation indices and texture from multispectral drone imagery can improve the accuracy of LAI estimation during the vegetative growth stage of winter wheat. In addition, selecting a flight height of 40 m can improve efficiency in large-scale agricultural field monitoring, as this study showed that drone data at flight heights of 20 m and 40 m did not significantly affect model accuracy.

Accuracy evaluation of reflectance, normalized difference vegetation index, and normalized difference water index using corrected unmanned aerial vehicle multispectral images by bidirectional reflectance distribution function and solar irradiance

Accuracy evaluation of reflectance, normalized difference vegetation index, and normalized difference water index using corrected unmanned aerial vehicle multispectral images by bidirectional reflectance distribution function and solar irradiance

Individualized Indicators and Estimation Methods for Tiger Nut (Cyperus esculentus L.) Tubers Yield Using Light Multispectral UAV and Lightweight CNN Structure

Tiger nuts are a non-genetically modified organism crop with high adaptability and economic value, and they are being widely promoted for cultivation in China. This study proposed a new yield-estimation method based on a lightweight convolutional neural network (CNN) named Squeeze Net to provide accurate production forecasts for tiger nut tubers. The multispectral unmanned aerial vehicle (UAV) images were used to establish phenotypic datasets of tiger nuts, comprising vegetation indices (VIs) and plant phenotypic indices. The Squeeze Net model with a lightweight CNN structure was constructed to fully explore the explanatory power of the spectral UAV-derived information and compare the differences between the parametric and nonparametric models applied in tiger nut yield predictions. Compared with stepwise multiple linear regression (SMLR), both algorithms achieved good yield prediction performances. The highest obtained accuracies reflected an R2 value of 0.775 and a root-mean-square error (RMSE) value of 688.356 kg/ha with SMLR, and R2 = 0.780 and RMSE = 716.625 kg/ha with Squeeze Net. This study demonstrated that Squeeze Net can efficiently process UAV multispectral images and improve the resolution and accuracy of the yield prediction results. Our study demonstrated the enormous potential of artificial intelligence (AI) algorithms in the precise crop management of tiger nuts in the arid sandy lands of northwest China by exploring the interactions between various intensive phenotypic traits and productivity.

Multimodal Deep Learning for Rice Yield Prediction Using UAV-Based Multispectral Imagery and Weather Data

Precise yield predictions are useful for implementing precision agriculture technologies and making better decisions in crop management. Convolutional neural networks (CNNs) have recently been used to predict crop yields in unmanned aerial vehicle (UAV)-based remote sensing studies, but weather data have not been considered in modeling. The aim of this study was to explore the potential of multimodal deep learning on rice yield prediction accuracy using UAV multispectral images at the heading stage, along with weather data. The effects of the CNN architectures, layer depths, and weather data integration methods on the prediction accuracy were evaluated. Overall, the multimodal deep learning model integrating UAV-based multispectral imagery and weather data had the potential to develop more precise rice yield predictions. The best models were those trained with weekly weather data. A simple CNN feature extractor for UAV-based multispectral image input data might be sufficient to predict crop yields accurately. However, the spatial patterns of the predicted yield maps differed from model to model, although the prediction accuracy was almost the same. The results indicated that not only the prediction accuracies, but also the robustness of within-field yield predictions, should be assessed in further studies.

Extraction and Mapping of Cropland Parcels in Typical Regions of Southern China Using Unmanned Aerial Vehicle Multispectral Images and Deep Learning

The accurate extraction of cropland distribution is an important issue for precision agriculture and food security worldwide. The complex characteristics in southern China pose great challenges to the extraction. In this study, for the objective of accurate extraction and mapping of cropland parcels in multiple crop growth stages in southern China, we explored a method based on unmanned aerial vehicle (UAV) data and deep learning algorithms. Our method considered cropland size, cultivation patterns, spectral characteristics, and the terrain of the study area. From two aspects—model architecture of deep learning and the data form of UAV—four groups of experiments are performed to explore the optimal method for the extraction of cropland parcels in southern China. The optimal result obtained in October 2021 demonstrated an overall accuracy (OA) of 95.9%, a Kappa coefficient of 89.2%, and an Intersection-over-Union (IoU) of 95.7%. The optimal method also showed remarkable results in the maps of cropland distribution in multiple crop growth stages, with an average OA of 96.9%, an average Kappa coefficient of 89.5%, and an average IoU of 96.7% in August, November, and December of the same year. This study provides a valuable reference for the extraction of cropland parcels in multiple crop growth stages in southern China or regions with similar characteristics.

Non-destructive monitoring of maize LAI by fusing UAV spectral and textural features.

Leaf area index (LAI) is an essential indicator for crop growth monitoring and yield prediction. Real-time, non-destructive, and accurate monitoring of crop LAI is of great significance for intelligent decision-making on crop fertilization, irrigation, as well as for predicting and warning grain productivity. This study aims to investigate the feasibility of using spectral and texture features from unmanned aerial vehicle (UAV) multispectral imagery combined with machine learning modeling methods to achieve maize LAI estimation. In this study, remote sensing monitoring of maize LAI was carried out based on a UAV high-throughput phenotyping platform using different varieties of maize as the research target. Firstly, the spectral parameters and texture features were extracted from the UAV multispectral images, and the Normalized Difference Texture Index (NDTI), Difference Texture Index (DTI) and Ratio Texture Index (RTI) were constructed by linear calculation of texture features. Then, the correlation between LAI and spectral parameters, texture features and texture indices were analyzed, and the image features with strong correlation were screened out. Finally, combined with machine learning method, LAI estimation models of different types of input variables were constructed, and the effect of image features combination on LAI estimation was evaluated. The results revealed that the vegetation indices based on the red (650 nm), red-edge (705 nm) and NIR (842 nm) bands had high correlation coefficients with LAI. The correlation between the linearly transformed texture features and LAI was significantly improved. Besides, machine learning models combining spectral and texture features have the best performance. Support Vector Machine (SVM) models of vegetation and texture indices are the best in terms of fit, stability and estimation accuracy (R2 = 0.813, RMSE = 0.297, RPD = 2.084). The results of this study were conducive to improving the efficiency of maize variety selection and provide some reference for UAV high-throughput phenotyping technology for fine crop management at the field plot scale. The results give evidence of the breeding efficiency of maize varieties and provide a certain reference for UAV high-throughput phenotypic technology in crop management at the field scale.

Nitrogen nutrition diagnosis for cotton under mulched drip irrigation using unmanned aerial vehicle multispectral images

Remote sensing has been increasingly used for precision nitrogen management to assess the plant nitrogen status in a spatial and real-time manner. The nitrogen nutrition index (NNI) can quantitatively describe the nitrogen status of crops. Nevertheless, the NNI diagnosis for cotton with unmanned aerial vehicle (UAV) multispectral images has not been evaluated yet. This study aimed to evaluate the performance of three machine learning models, i.e., support vector machine (SVM), back propagation neural network (BPNN), and extreme gradient boosting (XGB) for predicting canopy nitrogen weight and NNI of cotton over the whole growing season from UAV images. The results indicated that the models performed better when the top 15 vegetation indices were used as input variables based on their correlation ranking with nitrogen weight and NNI. The XGB model performed the best among the three models in predicting nitrogen weight. The prediction accuracy of nitrogen weight at the upper half-leaf level (R2=0.89, RMSE=0.68 g m–2, RE=14.62% for calibration and R2=0.83, RMSE=1.08 g m–2, RE=19.71% for validation) was much better than that at the all-leaf level (R2=0.73, RMSE=2.20 g m–2, RE=26.70% for calibration and R2=0.70, RMSE=2.48 g m–2, RE=31.49% for validation) and at the plant level (R2=0.66, RMSE=4.46 g m–2, RE=30.96% for calibration and R2=0.63, RMSE=3.69 g m–2, RE=24.81% for validation). Similarly, the XGB model (R2=0.65, RMSE=0.09, RE=8.59% for calibration and R2=0.63, RMSE=0.09, RE=8.87% for validation) also outperformed the SVM model (R2=0.62, RMSE=0.10, RE=7.92% for calibration and R2=0.60, RMSE=0.09, RE=8.03% for validation) and BPNN model (R2=0.64, RMSE=0.09, RE=9.24% for calibration and R2=0.62, RMSE=0.09, RE=8.38% for validation) in predicting NNI. The NNI predictive map generated from the optimal XGB model can intuitively diagnose the spatial distribution and dynamics of nitrogen nutrition in cotton fields, which can help farmers implement precise cotton nitrogen management in a timely and accurate manner.

Generating synthetic multispectral images using neural style transfer: A study with application in channel alignment

We can obtain information about food plantations by imaging technology such as satellite and unmanned aerial vehicles which enables rapid decision-making. Unmanned aerial vehicles provide fast and precise high-resolution multispectral imaging of plantation areas. However, unmanned aerial vehicle images are subject to some misalignment among their bands mainly due to the sensor’s displacement, vehicle movements, and others. To quantitatively evaluate channel alignment algorithms, it is necessary to have fully aligned images. For this purpose, we propose a method for generating fully aligned synthetic multispectral images based on real misaligned unmanned aerial vehicle multispectral images using Neural Style Transferring algorithms. We applied our methods over a real multispectral images dataset and we generated a new synthetic aligned image for each original misaligned multispectral image. Our experiments demonstrate that images generated by our method have fully aligned channels, and the new channels resemble the original ones. We also designed two case studies that demonstrate the applicability of our approach.

Improving the spatial and temporal estimation of ecosystem respiration using multi-source data and machine learning methods in a rainfed winter wheat cropland

The investigation of ecosystem respiration (RE) and its vital influential factors along with the timely and accurate detection of spatiotemporal variations in RE are essential for guiding agricultural production planning. RE observation in the plot region is primarily based on the laborious chamber method. However, upscaling the spatial-temporal estimates of RE at the canopy scale is still challenging. The present study conducted a field experiment to determine RE using the chamber method. A multi-rotor unmanned aerial vehicle (UAV) equipped with a multispectral camera was employed to acquire the canopy spectral data of wheat during each RE test experiment. Moreover, the agronomic indicators of aboveground plant biomass, leaf area index, leaf dry mass as well as agrometeorological and soil data were measured simultaneously. The study analyzed the potential of multi-information for estimating RE at the field scale and proposed two strategies for RE estimation. In addition, a semiempirical, yet Lloyd and Taylor-based, remote sensing model (LT1-NIRV) was developed for estimating RE observed across different growth stages with a small margin of error (coefficient of determination [R2] = 0.60–0.64, root-mean-square error [RMSE] = 285.98–316.19 mg m−2 h−1). Further, five machine learning (ML) algorithms were utilized to independently estimate RE using two different datasets. The rigorous analyses, which included statistical comparison and cross-validation for estimating RE, confirmed that the XGBoost model, with the highest R2 and lowest RMSE (R2 = 0.88 and RMSE = 172.70 mg m−2 h−1), performed the best among the evaluated ML models. The LT1-NIRV model was less effective in estimating RE compared with the other ML models. Based on this comprehensive comparison analysis, the ML model can successfully estimate variations in wheat field RE using high-resolution UAV multispectral images and environmental factors from the wheat cropland system, thereby providing a valuable reference for monitoring and upscaling RE observations.

Integrated diagnosis and time-series sensitivity evaluation of nutrient deficiencies in medicinal plant (Ligusticum chuanxiong Hort.) based on UAV multispectral sensors.

Nitrogen(N), phosphorus(P), and potassium(K) are essential elements that are highly deficient during plant growth. Existing diagnostic methods are not suitable for rapid diagnosis of large-scale planting areas. Near-ground remote sensing technology based on unmanned aerial vehicle (UAV) and sensor is often applied to crop growth condition monitoring and agricultural management. It has been proven to be used for monitoring plant N, P, and K content. However, its integrated diagnostic model has been less studied. In this study, we collected UAV multispectral images of Ligusticum chuanxiong Hort. in different periods of nutritional stress and constructed recognition models with different heights and algorithms. The optimal model variables were selected, and the effects of different sampling heights and modeling algorithms on the model efficiency under the time span were evaluated. At the same time, we evaluated the timeliness of the model based on leaf element content determination and SPAD. It was also validated in field crop production. The results showed that the LR algorithm's model had optimal performance at all periods and flight altitudes. The optimal accuracy of N-deficient plants identification reached 100%, P/K-deficient plants reached 92.4%, and normal plants reached 91.7%. The results of UAV multispectral diagnosis, chemical diagnosis, and SPAD value diagnosis were consistent in the diagnosis of N deficiency, and the diagnosis of P and K deficiency was slightly lagging behind that of chemical diagnosis. This research uses UAV remote sensing technology to establish an efficient, fast, and timely nutritional diagnosis method for L. Chuanxiong, which is applied in production. Meanwhile, the standardized production of medicinal plant resources provides new solutions.

A Back Propagation Neural Network-Based Radiometric Correction Method (BPNNRCM) for UAV Multispectral Image

Radiometric correction is one of the most important preprocessing parts of unmanned aerial vehicle (UAV) multispectral remote sensing data analysis and application. In this article, a back propagation (BP) neural network-based radiometric correction method (BPNNRCM) considering optimal parameters was proposed. First, we used different UAV multispectral sensors (K6 equipped on the DJI M600, D-MSPC2000 equipped on the FEIMA D2000) to collect training, validation, testing and cross-validation data. Second, the radiometric correction results of BP neural network with different input variables and hidden layer node number were compared to select the best combination of input parameters and hidden layer node number. Finally, the radiometric correction accuracy and robustness of BP neural network considering the optimal parameters were verified. When the number of nodes in the input layer was five (digital number, UAV sensor height, wavelength, solar altitude angle, and temperature) and the number of nodes in the hidden layer was eight, the BP neural network had the best comprehensive performance in training time of train set and accuracy of validation/test set. In the aspect of accuracy and robustness, the absolute errors of test and cross-validation images' surface reflectance obtained by the BPNNRCM were all less than 0.054. The BPNNRCM had smaller average absolute error (0.0141), mean squared error (0.0003), mean absolute error (0.0141) and mean relative error (7.1%) comparing with empirical line method and radiative transfer model. In general, the research results of this article prove the feasibility and prospect of BPNNRCM for radiometric correction of UAV multispectral images.

Estimating maize LAI by exploring deep features of vegetation index map from UAV multispectral images

Visible and near-infrared imaging spectroscopy is an efficient method for non-destructive estimation of crop leaf area index (LAI), in which effective features extraction is a key factor affecting model accuracy. Vegetation index like normalized difference vegetation index (NDVI) is a typical parameter extracted from spectral images for LAI estimation. However, it always faces challenges of values saturation and sensitivity decreasing due to the canopy coverage changes and complex environment influences. Therefore, the study aims to propose a method to explore deep features based on deep learning model (ResNet50) and vegetation index (VI) map, by which contributes to improve the accuracy of maize LAI estimation. In two-years’ experiments of 2020 and 2021, the multi-spectral imaging sensor carried by the unmanned aerial vehicle (UAV) was used to collected remote sensing images of the maize canopy during multi-growth stages under different nitrogen fertilizer treatments, and a total of 792 LAI values were collected. In terms of feature extraction, 10 VIs were extracted based on the spectral images of maize canopy, and comprehensive features (texture and deep features) were extracted based on the VI map. In terms of model construction, partial least squares regression (PLS) models were constructed based on three feature types. The results showed that normalized difference red edge (NDRE) VI was the best indicator for monitoring maize LAI among all the VIs investigated. Meanwhile, LAI multivariate linear model constructed based on texture features extracted from NDRE images had better accuracy. Deep features of NDRE images were further extracted by ResNet50 model. And the maize LAI estimation models based on PLS were constructed with VI, texture features and deep features of two-year data, respectively. The results showed that compared with VI and texture features, the LAI estimation model based on deep features showed optimal accuracy in two-year data ( R 2 = 0.827, greater than 0.781 and 0.780; RMSEP = 0.405, less than 0.455 and 0.455; MAE = 0.320, less than 0.366 and 0.363). In addition, the LAI estimation model based on deep features was successfully evaluated with 2020 and 2021 validation data (2020: R 2 = 0.872, RMSEP = 0.379, MAE = 0.294; 2021: R 2 = 0.733, RMSEP = 0.433, MAE = 0.348). The experimental results demonstrate the effectiveness of the proposed deep feature extraction method based on deep learning model and VI images in LAI estimation, which provides a feasible method for monitoring crop growth information based on UAV platform.

Comparison of PROSAIL Model Inversion Methods for Estimating Leaf Chlorophyll Content and LAI Using UAV Imagery for Hemp Phenotyping

Unmanned aerial vehicle (UAV) remote sensing was used to estimate the leaf area index (LAI) and leaf chlorophyll content (LCC) of two hemp cultivars during two growing seasons under four nitrogen fertilisation levels. The hemp traits were estimated by the inversion of the PROSAIL model from UAV multispectral images. The look-up table (LUT) and hybrid regression inversion methods were compared. The hybrid methods performed better than LUT methods, both for LAI and LCC, and the best accuracies were achieved by random forest for the LAI (0.75 m2 m−2 of RMSE) and by Gaussian process regression for the LCC (9.69 µg cm−2 of RMSE). High-throughput phenotyping was carried out by applying a generalised additive model to the time series of traits estimated by the PROSAIL model. Through this approach, significant differences in LAI and LCC dynamics were observed between the two hemp cultivars and between different levels of nitrogen fertilisation.

Remote estimation of leaf nitrogen concentration in winter oilseed rape across growth stages and seasons by correcting for the canopy structural effect

Estimating crop leaf nitrogen concentration (LNC, %) with the canopy bidirectional reflectance factor (BRF) is an effective method for detecting the nitrogen (N) deficiency in crops. It is challenging to remotely estimate LNC across growth stages and seasons with a general empirical model, since the complex change in canopy structure under N deficiencies and across growth stages affects the accuracy of the estimations. The canopy scattering coefficient (CSC), the ratio of the BRF to the directional area scattering factor (DASF), has been suggested to reduce the canopy structural effect on BRF. However, the DASF can only be calculated for closed canopies and is not applicable to the early growth stages of crops when the fields are sparsely vegetated. This study proposed a new method for decoupling the canopy structural effect and canopy BRF using the near-infrared reflectance of vegetation (NIRV). NIRV is driven by the change in canopy structure while mitigates the soil contribution. The method was tested through six field experiments on ten farmers' fields of winter oilseed rape (Brassica napus L.) using both in-situ hyperspectral data and unmanned aerial vehicle (UAV) multispectral images. The results demonstrated that NIRV was closely related to the leaf area index (LAI) (R2 = 0.79) across growth stages and seasons. The CSC was derived with NIRV based on the linear relationship between NIRV and DASF for the closed canopies. The LNC predicted by the NIRV-derived CSC (R2 = 0.69, RMSE = 0.51 for in-situ hyperspectral data and R2 = 0.65, RMSE = 0.49 for UAV multispectral images) was more accurate than the results derived from the BRF (R2 = 0.55, RMSE = 0.62 for in-situ hyperspectral data and R2 = 0.59, RMSE = 0.60 for UAV multispectral images) with the independent dataset, suggesting that correcting for the canopy structural effect with NIRV provided a new alternative for suppressing the impact of canopy structure on the canopy BRF. As NIRV is easily calculated with diverse remote sensing data sources, this study proved the potential of applying NIRV to improving the accuracy and transferability of the LNC prediction model at different scales.