Hyperspectral Sensors Research Articles

Precision Soil Moisture Monitoring Through Drone-Based Hyperspectral Imaging and PCA-Driven Machine Learning

Accurately estimating soil moisture at multiple depths is essential for sustainable farming practices, as it supports efficient irrigation management, optimizes crop yields, and conserves water resources. This study integrates a drone-mounted hyperspectral sensor with machine learning techniques to enhance soil moisture estimation at 10 cm and 30 cm depths in a cornfield. The primary aim was to understand the relationship between root zone water content and canopy reflectance, pinpoint the depths where this relationship is most significant, identify the most informative wavelengths, and train a machine learning model using those wavelengths to estimate soil moisture. Our results demonstrate that PCA effectively detected critical variables for soil moisture estimation, with the ANN model outperforming other machine learning algorithms, including Random Forest (RF), Support Vector Regression (SVR), and Gradient Boosting (XGBoost). Model comparisons between irrigated and non-irrigated treatments showed that soil moisture in non-irrigated plots could be estimated with greater accuracy across various dates. This finding indicates that plants experiencing high water stress exhibit more significant spectral variability in their canopy, enhancing the correlation with soil moisture in the root zone. Moreover, over the growing season, when corn exhibits high chlorophyll content and increased resilience to environmental stressors, the correlation between canopy spectrum and root zone soil moisture weakens. Error analysis revealed the lowest relative estimation errors in non-irrigated plots at a 30 cm depth, aligning with periods of elevated water stress at shallower levels, which drove deeper root growth and strengthened the canopy reflectance relationship. This correlation corresponded to lower RMSE values, highlighting improved model accuracy.

Spectral unmixing of hyperspectral images revealed pine wilt disease sensitive endmembers.

Throughout the entire cycle of leaf phenological events, leaf colour undergoes changes that are influenced by either abiotic stress or biotic infection. These changes in colouration are closely linked to the quantity and quality of photosynthetic pigments, which directly impact the primary productivity of plants. Therefore, monitoring and quantifying leaf colouration changes are crucial for distinguishing damage caused by pine wilt nematodes from natural tree senescence. In this study, a hyperspectral camera sensor was employed for the non-invasive and non-destructive evaluation of needle colour changes in coniferous trees grown in field tests. Three distinct needle colour variations of six coniferous tree species were selected and monitored using a hyperspectral sensor: those displaying seasonal autumn colours, undergoing nematode-infected necrosis processes, and experiencing natural death. To mitigate the inherently mixed spectral properties of hyperspectral data, endmembers were extracted from individual images using the Purity Pixel Index algorithm under the assumption of linear mixing of endmembers. From a total of 1,321 endmembers extracted from 378 hyperspectral images of six pine species, eight endmembers were ultimately chosen to reconstruct hyperspectral images and generate abundance maps. Among these eight endmembers, four represent varying levels of photosynthetic pigment contents-ranging from very low to high. Consequently, these coniferous endmembers hold promise for assessing seasonal leaf phenology and the extent of damage in pine trees infected by pine wilt nematodes. This comprehensive approach underscores the effectiveness of spectral unmixing of hyperspectral images in advancing precision forestry through meticulous coniferous needle trait analysis.

Integration of PRISMA Hyperspectral Satellite Data with Ground Based Geological investigation for Mapping alteration minerals associated with the Neem-Ka-Thana Cu belt in Rajasthan, India

Integration of PRISMA Hyperspectral Satellite Data with Ground Based Geological investigation for Mapping alteration minerals associated with the Neem-Ka-Thana Cu belt in Rajasthan, India

Espectroscopia VIS-NIR-SWIR na cultura da cana-de-açúcar (Saccharum officinarum L.) para fins fitossanitários

ABSTRACT: One of the causes of productivity loss in sugarcane cultivation has been associated with the species Diatraea saccharalis, also known as the sugarcane borer. Therefore, this study evaluated the feasibility of using hyperspectral sensors to obtain the leaf spectral response of sugarcane in different periods of infestation at leaf and canopy levels to diagnose D. saccharalis damage in advance. The study included three varieties of sugarcane: CTC9003BT, CTC4, and RB966928. The insecticide Altacor® was used to control pestsin half of the plots. Data collection occurred at the following stages: sprouting, tillering, stalke longation, and early maturation. Data regarding relative water content (RWC), leaf spectral signature in the laboratory and canopy, vegetation indices (NDVI and MCARI), productivity, purity, and total recoverable sugar (TRS) were collected to evaluate the borer effects on the sugarcane crop. The highest RWC was observed for CTC9003BT (60.15%) without insecticide and 59.0% with insecticide. The visible (400-680 nm) and near-infrared (750-1300 nm) spectral bands identified spectral variations in plants with and without sugarcane borer. The percentage of sugarcane borer showed high and negative correlations between productivity, NDVI, and TRS, with coefficients of -0.68, -0.76, and -0.76, respectively. The NDVI and MCARI indices effectively detect plants under stress, but their variation is influenced by multiple factors, making it difficult to associate them with a single problem (sugarcane borer).

Drone-Based Sensing and Imaging for Fruit Crop Monitoring: A Review

The global fruit production industry has been suffering from numerous problems in the yield, quality, and food safety context. The use of drone-based sensing and imaging technologies has emerged as a promising approach for monitoring fruit crops, enabling real-time assessment of crop health, growth, and development. Monitoring fruit crops helps identify areas of improvement and makes decisions based on data. This review focuses on the various sensor technologies utilized in drone-based fruit crop monitoring, including RGB, multispectral, hyperspectral, thermal, and LiDAR sensors. The applications of these sensors are discussed, including yield estimation and prediction, crop growth monitoring, disease detection and diagnosis, pest detection and management, nutrient deficiency detection, and water stress monitoring. The review highlights the advantages and limitations of each sensor technology, as well as the challenges associated with data processing and analysis. Case studies demonstrate the effectiveness of drone-based sensing and imaging in fruit crop monitoring, and future directions are discussed, including the integration of sensor technologies with other precision agriculture tools and the development of specialized sensors and cameras. Standardization and best practices are emphasized as crucial for the widespread adoption of drone-based sensing and imaging in fruit crop monitoring.

Effects of functional type and angular configuration on reflectance anisotropy of aquatic vegetation in ultra-high resolution hyperspectral imagery

ABSTRACT Imaging spectroscopy and lightweight unmanned aerial vehicles (UAVs) have revolutionized remote sensing of vegetation, by providing high spectral and spatial resolution data. Comprehensive wetland monitoring should be based on reliable mapping of biophysical aquatic vegetation parameters, which in turn requires low bias in measured spectra, e.g. minimization of reflectance anisotropy. This study aims to quantitatively investigate how sensor scan direction and solar zenith angle (SZA) affect vegetation spectra over two dominant aquatic plants, Phragmites australis and Nuphar lutea, representing different functional types (i.e. emergent and floating) with divergent canopy structure and affinity with water, using data collected with a hyperspectral (400–1000 nm range) push-broom sensor mounted on a UAV. Anisotropy factor (ANIF), mean reflectance scatterplots and regression analysis were used for assessing the magnitude of spectral anisotropy at both pseudo-leaf and canopy scales. Our findings showed more accentuated anisotropic behaviour in the visible than near-infrared domain at SZA < 60° as the scan direction approaches that of solar principal plane. The increase in reflectance in near-forward direction affects the spectra of floating-leaved N. lutea at both leaf and canopy scales, suggesting the presence of water (as a film over leaves or as canopy background) as a likely anisotropy driver. Backward hotspot is evident in canopy spectra of emergent P. australis, with driving mechanisms similar to terrestrial species (i.e. shadow-hiding). At SZA > 60°, reflectance anisotropy appears to be negligible, independent of functional type and scan direction. This research underscores the importance of considering sun-target-sensor geometry in remote sensing studies on aquatic vegetation.

Target Detection and Characterization of Multi-Platform Remote Sensing Data

Detecting targets in remote sensing imagery, particularly when identifying sparsely distributed materials, is crucial for applications such as defense, mineral exploration, agriculture, and environmental monitoring. The effectiveness of detection and the precision of the results are influenced by several factors, including sensor configurations, platform properties, interactions between targets and their background, and the spectral contrast of the targets. Environmental factors, such as atmospheric conditions, also play a significant role. Conventionally, target detection in remote sensing has relied on statistical methods that typically assume a linear process for image formation. However, to enhance detection performance, it is critical to account for the geometric and spectral variabilities across multiple imaging platforms. In this research, we conducted a comprehensive target detection experiment using a unique benchmark multi-platform hyperspectral dataset, where man-made targets were deployed on various surface backgrounds. Data were collected using a hand-held spectroradiometer, UAV-mounted hyperspectral sensors, and airborne platforms, all within a half-hour time window. Multi-spectral space-based sensors (i.e., Worldview and Landsat) also flew over the scene and collected data. The experiment took place on 23 July 2021, at the Rochester Institute of Technology’s Tait Preserve in Penfield, NY, USA. We validated the detection outcomes through receiver operating characteristic (ROC) curves and spectral similarity metrics across various detection algorithms and imaging platforms. This multi-platform analysis provides critical insights into the challenges of hyperspectral target detection in complex, real-world landscapes, demonstrating the influence of platform variability on detection performance and the necessity for robust algorithmic approaches in multi-source data integration.

Differentiation of Soybean Genotypes Concerning Seed Physiological Quality Using Hyperspectral Bands

The use of summarized spectral data in bands obtained by hyperspectral sensors can make it possible to obtain biochemical information about seeds and, thus, relate the results to seed viability and vigor. Thus, the hypothesis of this work is based on the possibility of obtaining information about the physiological quality of seeds through hyperspectral bands and distinguishing seed lots regarding their quality through wavelengths. The objective was then to evaluate the possibility of differentiating soybean genotypes regarding the physiological quality of seeds using spectral data. The experiment was conducted during the 2021/2022 harvest at the Federal University of Mato Grosso do Sul in a randomized block design with four replicates and 10 F3 soybean populations (G1, G8, G12, G15, G19, G21, G24, G27, G31, and G36). After the maturation of each genotype, seeds were harvested from the central rows of each plot, which consisted of five one-meter rows. Seed samples from each experimental unit were placed in a Petri dish to collect spectral data. Readings were performed in the laboratory at a temperature of 26 °C and using two 60 W halogen lamps as the light source, positioned 15 cm between the sensor and the sample. The sensor used was the Ocean Optics (Florida, USA) model STS-VIS-L-50-400-SMA, which captured the reflectance of the seed sample at wavelengths between 450 and 824 nm. After readings from the hyperspectral sensor, the seeds were subjected to tests for water content, germination, first germination count, electrical conductivity, and tetrazolium. The data obtained were subjected to an analysis of variance and the means were compared by the Scott–Knott test at 5% probability, analyzed using R software version 4.2.3 (Auckland, New Zealand). The data on the physiological quality of the seeds of the soybean genotypes were subjected to principal component analysis (PCA) and associated with the K-means algorithm to form groups according to the similarity and distinction between the genetic materials. After the formation of these groups, spectral curve graphs were constructed for each soybean genotype and for the groups that were formed. The physiological quality of the soybean genotypes can be differentiated using hyperspectral bands. The spectral bands, therefore, provide important information about the physiological quality of soybean seeds. Through the use of hyperspectral sensors and the observation of specific bands, it is possible to differentiate genotypes in terms of seed quality, complementing and/or replacing traditional tests in a fast, accurate, and non-destructive way, reducing the time and investment spent on obtaining information on seed viability and vigor. The results found in this study are promising, and further research is needed in future studies with other species and genotypes. The interval between 450 and 649 nm was the main spectrum band that contributed to the differentiation between soybean genotypes of superior and inferior physiological quality.

Detection of Irrigated and Non-Irrigated Soybeans Using Hyperspectral Data in Machine-Learning Models

The objectives of this work are (i) to classify soybean cultivars under different irrigation managements using hyperspectral data, looking for the best machine-learning algorithm for the classification and the input that improves the performance of the models. The experiment was implemented in the 2023/24 harvest in the experimental area of the Federal University of Mato Grosso do Sul, Câmpus Chapadão do Sul, Mato Grosso do Sul, and it was conducted in a strip scheme with seven cultivars subjected to irrigated and rainfed management. Sixty days after crop emergence, three leaves per plot were collected for evaluation by the hyperspectral sensor. The spectral data was then separated into 28 bands to reduce dimensionality. In this way, two databases were generated: one with all the spectral information provided by the sensor (WL) and one with the 28 spectral bands (SB). Each database was subjected to different machine-learning models to ascertain the improved accuracy of the models in distinguishing the different eucalyptus species. The models tested were artificial neural networks (ANN), decision trees (DT), linear regression (LR), M5P algorithm, random forest (RF), and support vector machine (SVM). The results demonstrate the effectiveness of machine-learning models in differentiating soybean management under rainfed and irrigated conditions, highlighting the advantage of hyperspectral data (WL) over selected spectral bands (SB). Models such as the support vector machine (SVM) showed the best levels of accuracy when using the entire available spectrum. On the other hand, artificial neural networks (ANN) performed well with spectral band data, demonstrating their ability to work with smaller data sets without compromising the classification.

Stacked Ensemble with Machine Learning Regressors on Optimal Features (SMOF) of hyperspectral sensor PRISMA for inland water turbidity prediction.

Leveraging hyperspectral data across various domains yields substantial benefits, yet managing many spectral bands and identifying the essential ones poses a formidable challenge. This study identifies the most relevant bands within a hyperspectral data cube for turbidity prediction in inland water. Nine machine learning regressors Cat Boost, Decision Trees, Extra Trees, Gradient Boost, Light Gradient Boost (LightGBM), Recursive Feature Elimination (RFE), Random Forest, Support Vector Regressor (SVR), and Xtreme Gradient Boost (XGBoost) have been used to compute the feature importance of the hyperspectral bands for predicting turbidity. Random Forest has outperformed the other models with a mean absolute percentage error (MAPE) of 1.61%, and the R2 of the linear fit is 0.96. Band 77, with a central wavelength of 1067.61nm, is the most dominating band regarding feature importance. We have also developed a novel framework for turbidity prediction: Stacked Ensemble with Machine Learning Regressors on Optimal Features (SMOF). It employs a stacking ensemble of the nine regressors mentioned above with Random Forest as both base and meta-model, leveraging feature selection outputs. With this framework, the MAPE (%) reached 1.21, while the R2 stood at 0.95. The present study also presents a simple statistical algorithm to detect noisy bands in the Hyperspectral Precursor of the Application Mission (PRISMA) data cube. The approach assesses quadrat-wise intra-band spatial coherence using Renyi's entropy thresholding for noisy band segregation. Radiometric calibration error and absorption due to water vapour are the two primary sources of noise within the data cube. Moreover, this research implements the open-source Water Colour Simulator (WASI) to simulate inland water spectra with varied proportions of turbidity. Overall, the study presents an approach to identify noisy bands and integrates the potential wavelengths for turbidity prediction of inland waters.

Remote Sensing Technologies Using UAVs for Pest and Disease Monitoring: A Review Centered on Date Palm Trees

This review is aimed at exploring the use of remote sensing technology with a focus on Unmanned Aerial Vehicles (UAVs) in monitoring and management of palm pests and diseases with a special focus on date palms. It highlights the most common sensor types, ranging from passive sensors such as RGB, multispectral, hyperspectral, and thermal as well as active sensors such as light detection and ranging (LiDAR), expounding on their unique functions and gains as far as the detection of pest infestation and disease symptoms is concerned. Indices derived from UAV multispectral and hyperspectral sensors are used to assess their usefulness in vegetation health monitoring and plant physiological changes. Other UAVs are equipped with thermal sensors to identify water stress and temperature anomalies associated with the presence of pests and diseases. Furthermore, the review discusses how LiDAR technology can be used to capture detailed 3D canopy structures as well as volume changes that may occur during the progressing stages of a date palm infection. Besides, the paper examines how machine learning algorithms have been incorporated into remote sensing technologies to ensure high accuracy levels in detecting diseases or pests. This paper aims to present a comprehensive outline for future research focusing on modern methodologies, technological improvements, and direction for the efficient application of UAV-based remote sensing in managing palm tree pests and diseases.

Drones in Precision Agriculture: A Comprehensive Review of Applications, Technologies, and Challenges

In the face of growing challenges in modern agriculture, such as climate change, sustainable resource management, and food security, drones are emerging as essential tools for transforming precision agriculture. This systematic review, based on an in-depth analysis of recent scientific literature (2020–2024), provides a comprehensive synthesis of current drone applications in the agricultural sector, primarily focusing on studies from this period while including a few notable exceptions of particular interest. Our study examines in detail the technological advancements in drone systems, including innovative aerial platforms, cutting-edge multispectral and hyperspectral sensors, and advanced navigation and communication systems. We analyze diagnostic applications, such as crop monitoring and multispectral mapping, as well as interventional applications like precision spraying and drone-assisted seeding. The integration of artificial intelligence and IoTs in analyzing drone-collected data is highlighted, demonstrating significant improvements in early disease detection, yield estimation, and irrigation management. Specific case studies illustrate the effectiveness of drones in various crops, from viticulture to cereal cultivation. Despite these advancements, we identify several obstacles to widespread drone adoption, including regulatory, technological, and socio-economic challenges. This study particularly emphasizes the need to harmonize regulations on beyond visual line of sight (BVLOS) flights and improve economic accessibility for small-scale farmers. This review also identifies key opportunities for future research, including the use of drone swarms, improved energy autonomy, and the development of more sophisticated decision-support systems integrating drone data. In conclusion, we underscore the transformative potential of drones as a key technology for more sustainable, productive, and resilient agriculture in the face of global challenges in the 21st century, while highlighting the need for an integrated approach combining technological innovation, adapted policies, and farmer training.

Transforming Farming: A Review of AI-Powered UAV Technologies in Precision Agriculture

The integration of unmanned aerial vehicles (UAVs) with artificial intelligence (AI) and machine learning (ML) has fundamentally transformed precision agriculture by enhancing efficiency, sustainability, and data-driven decision making. In this paper, we present a comprehensive overview of the integration of multispectral, hyperspectral, and thermal sensors mounted on drones with AI-driven algorithms to transform modern farms. Such technologies support crop health monitoring in real time, resource management, and automated decision making, thus improving productivity with considerably reduced resource consumption. However, limitations include high costs of operation, limited UAV battery life, and the need for highly trained operators. The novelty of this study lies in the thorough analysis and comparison of all UAV-AI integration research, along with an overview of existing related works and an analysis of the gaps. Furthermore, practical solutions to technological challenges are summarized to provide insights into precision agriculture. This paper also discusses the barriers to UAV adoption and suggests practical solutions to overcome existing limitations. Finally, this paper outlines future research directions, which will discuss advances in sensor technology, energy-efficient AI models, and how these aspects influence ethical considerations regarding the use of UAVs in agricultural research.

Raman spectroscopy combined with active hyperspectral sensing for classification of waste plastics containing brominated flame retardants: A sensor fusion approach.

Discrimination of waste plastics according to brominated flame retardant (BFR) concentration is essential to ensure quality and safety in recycling. We present a sensor fusion approach to classify BFR-containing plastic waste by combining Raman and near-infrared (NIR) spectroscopies. We analysed 210 waste plastic samples sourced from waste electronics and electrical equipment stream and 25 laboratory-made plastics. The Raman spectra were acquired in the range 27-2481 cm-1 using a time-gated Raman and the NIR spectra in the range 4000-5260 cm-1 using a novel active hyperspectral sensor. Total elemental bromine concentrations were determined with X-ray fluorescence spectroscopy and used as reference for training extremely randomized trees classifiers for high- and low-bromine plastics with different thresholds of segmentation. The classifier models were built using Raman and NIR spectral data after reducing dimensions with principal component analysis, both separately and by fusing the data. We achieved over 80% balanced classification accuracies using all models, with significant improvements by data fusion.

A broadband hyperspectral image sensor with high spatio-temporal resolution

Hyperspectral imaging provides high-dimensional spatial–temporal–spectral information showing intrinsic matter characteristics1–5. Here we report an on-chip computational hyperspectral imaging framework with high spatial and temporal resolution. By integrating different broadband modulation materials on the image sensor chip, the target spectral information is non-uniformly and intrinsically coupled to each pixel with high light throughput. Using intelligent reconstruction algorithms, multi-channel images can be recovered from each frame, realizing real-time hyperspectral imaging. Following this framework, we fabricated a broadband visible–near-infrared (400–1,700 nm) hyperspectral image sensor using photolithography, with an average light throughput of 74.8% and 96 wavelength channels. The demonstrated resolution is 1,024 × 1,024 pixels at 124 fps. We demonstrated its wide applications, including chlorophyll and sugar quantification for intelligent agriculture, blood oxygen and water quality monitoring for human health, textile classification and apple bruise detection for industrial automation, and remote lunar detection for astronomy. The integrated hyperspectral image sensor weighs only tens of grams and can be assembled on various resource-limited platforms or equipped with off-the-shelf optical systems. The technique transforms the challenge of high-dimensional imaging from a high-cost manufacturing and cumbersome system to one that is solvable through on-chip compression and agile computation.

Towards the Future of Ubiquitous Hyperspectral Imaging: Innovations in Sensor Configurations and Cost Reduction for Widespread Applicability

Hyperspectral imaging is currently under active development as a method for remote sensing, environmental monitoring and biomedical diagnostics. The development of hyperspectral sensors is aimed at their miniaturization and reducing the cost of components for the purpose of the widespread use of such devices on unmanned aerial vehicles and satellites. In this review, we present a broad overview of recent work on the development of hyperspectral devices’ configurations, studies aimed at modifying sensors and the possibility of reducing the cost of components of such devices. In addition, we will present the main trends in the development of hyperspectral device configurations for ubiquitous applications.

'Candidatus Liberibacter asiaticus' infection alters the reflectance profile in asymptomatic citrus plants.

Huanglongbing (HLB) is the primary and most destructive disease affecting citrus, caused by a pathogen transmitted by an insect vector, Diaphorina citri. There are no curative methods for the disease, and rapid and accurate methods are needed for early detection in the field, even before symptoms appear. These will facilitate the faster removal of infected trees, preventing the spread of the bacteria through commercial citrus orchards. It was possible to determine ranges of hyperspectral bands that demonstrated significant differences in relative reflectance between treatments consisting of healthy and infected plants from the first days of evaluation, when plants infected with 'Candidatus Liberibacter asiaticus' (CLas) were still in the asymptomatic stage of the disease. From the Week 2 of evaluation [58 days after infection (DAI) of plants] until the last week, spectral differences were detected in the red edge region (660-750 nm). From the Week 6 onwards (86 DAI), spectral differences between healthy and symptomatic plants were observed in bands close to the visible region (520-680 nm). Spectral differences were detected in the leaves of C. sinensis infected by CLas before the appearance of symptoms, making it feasible to use the hyperspectral sensor to monitor the disease. Our results indicate the need for future studies to validate the use of hyperspectral sensors for managing and detecting HLB in commercial citrus orchards, contributing to the integrated management of the disease. © 2024 Society of Chemical Industry.

Large-Kernel Central Block Masked Convolution and Channel Attention-Based Reconstruction Network for Anomaly Detection of High-Resolution Hyperspectral Imagery

In recent years, the rapid advancement of drone technology has led to an increasing use of drones equipped with hyperspectral sensors for ground imaging. Hyperspectral data captured via drones offer significantly higher spatial resolution, but this also introduces more complex background details and larger target scales in high-resolution hyperspectral imagery (HRHSI), posing substantial challenges for hyperspectral anomaly detection (HAD). Mainstream reconstruction-based deep learning methods predominantly emphasize spatial local information in hyperspectral images (HSIs), relying on small spatial neighborhoods for reconstruction. As a result, large anomalous targets and background details are often well reconstructed, leading to poor anomaly detection performance, as these targets are not sufficiently distinguished from the background. To address these limitations, we propose a novel HAD network for HRHSI based on large-kernel central block masked convolution and channel attention, termed LKCMCA. Specifically, we first employ the pixel-shuffle technique to reduce the size of anomalous targets without losing image information. Next, we design a large-kernel central block masked convolution to make the network pay more attention to the surrounding background information, enabling better fusion of the information between adjacent bands. This, coupled with an efficient channel attention mechanism, allows the network to capture deeper spectral features, enhancing the reconstruction of the background while suppressing anomalous targets. Furthermore, we introduce an adaptive loss function by down-weighting anomalous pixels based on the mean absolute error. This loss function is specifically designed to suppress the reconstruction of potentially anomalous pixels during network training, allowing our model to be considered an excellent background reconstruction network. By leveraging reconstruction error, the model effectively highlights anomalous targets. Meanwhile, we produced four benchmark datasets specifically for HAD tasks using existing HRHSI data, addressing the current shortage of HRHSI datasets in the HAD field. Extensive experiments demonstrate that our LKCMCA method achieves superior detection performance, outperforming ten state-of-the-art HAD methods on all datasets.

Machine Learning and New-Generation Spaceborne Hyperspectral Data Advance Crop Type Mapping

Hyperspectral sensors provide near-continuous spectral data that can facilitate advancements in agricultural crop classification and characterization, which are important for addressing global food and water security issues. We investigated two new-generation hyperspectral sensors, Germany’s Deutsches Zentrum für Luft‐ und Raumfahrt Earth Sensing Imaging Spectrometer (DESIS) and Italy’s PRecursore IperSpettrale della Missione Applicativa (PRISMA), within California???s Central Valley in August 2021 focusing on five irrigated agricultural crops (alfalfa, almonds, corn, grapes, and pistachios). With reference data from the U.S. Department of Agriculture Cropland Data Layer, we developed a spectral library of the crops and classified them using three machine learning algorithms (support vector machines [SVM], random forest [RF], and spectral angle mapper [SAM]) and two philosophies: 1. Full spectral analysis (FSA) and 2. Optimal hyperspectral narrowband (OHNB) analysis. For FSA, we used 59 DESIS four-bin product bands and 207 of 238 PRISMA bands. For OHNB analysis, 9 DESIS and 16 PRISMA nonredundant OHNBs for studying crops were selected. FSA achieved only 1% to 3% higher accuracies relative to OHNB analysis in most cases. SVM provided the best results, closely followed by RF. Using both DESIS and PRISMA image OHNBs in SVM for classification led to higher accuracy than using either image alone, with an overall accuracy of 99%, producer’s accuracies of 94% to 100%, and user???s accuracies of 95% to 100%.

Mapping Leaf Mass Per Area and Equivalent Water Thickness from PRISMA and EnMAP

With the continued advancement of spaceborne hyperspectral sensors, hyperspectral remote sensing is evolving as an increasingly pivotal tool for high-precision global monitoring applications. Novel image spectroscopy data, e.g., the PRecursore IperSpettrale della Missione Applicativa (PRISMA) and Environmental Mapping and Analysis Program (EnMAP), can rapidly and non-invasively capture subtle spectral information of terrestrial vegetation, facilitating the precise retrieval of the required vegetation parameters. As critical vegetation traits, Leaf Mass per Area (LMA) and Equivalent Water Thickness (EWT) hold significant importance for comprehending ecosystem functionality and the physiological status of plants. To address the demand for high-precision vegetation parameter datasets, a hybrid modeling approach was proposed in this study, integrating the radiative transfer model PROSAIL and neural network models to retrieve LMA and EWT from PRISMA and EnMAP images. To achieve this objective, canopy reflectance was simulated via PROSAIL, and the optimal band combinations for LMA and EWT were selected as inputs to train neural networks. The evaluation of the hybrid inversion models over field measurements showed that the RMSE values for the LMA and EWT were 4.11 mg·cm−2 and 9.08 mg·cm−2, respectively. The hybrid models were applied to PRISMA and EnMAP images, resulting in LMA and EWT maps displaying adequate spatial consistency, along with cross-validation results showing high accuracy (RMSELMA = 5.78 mg·cm−2, RMSEEWT = 6.84 mg·cm−2). The results demonstrated the hybrid inversion model’s universality and applicability, enabling the retrieval of vegetation parameters from image spectroscopy data and offering a valuable contribution to hyperspectral remote sensing for vegetation monitoring, though the availability of field measurement data remained a significant challenge.