Search Results (57)

Developing an effective vegetation classification method for mountain–plain transition zones is critical for understanding ecological patterns, evaluating ecosystem services, and guiding conservation efforts. Existing methods perform well in mountainous and plain areas but lack verification in mountain–plain transition zones. This study utilized terrain data and Sentinel-1 and Sentinel-2 imagery to extract topographic, spectral, texture, and SAR features as well as the vegetation index. By combining feature sets and applying feature elimination algorithms, the classification performance of one-dimensional convolutional neural networks (1D-CNNs), Random Forest (RF), and Multilayer Perceptron (MLP) was evaluated to determine the optimal feature combinations and methods. The results show the following: (1) multi-feature combinations, especially spectral and topographic features, significantly improved classification accuracy; (2) Recursive Feature Elimination based on Random Forest (RF-RFE) outperformed ReliefF in feature selection, identifying more representative features; (3) all three algorithms performed well, with consistent spatial results. The MLP algorithm achieved the best overall accuracy (OA: 81.65%, Kappa: 77.75%), demonstrating robustness and lower dependence on feature quantity. This study presents an efficient and robust vegetation classification workflow, verifies its applicability in mountain–plain transition zones, and provides valuable insights for small-region vegetation classification under similar topographic conditions globally. Full article

Backgrounds: Virtual reality (VR) has become a transformative technology with applications in gaming, education, healthcare, and psychotherapy. The subjective experiences in VR vary based on the virtual environment’s characteristics, and electroencephalography (EEG) is instrumental in assessing these differences. By analyzing EEG signals, researchers can explore the neural mechanisms underlying cognitive and emotional responses to VR stimuli. However, distinguishing EEG signals recorded by two-dimensional (2D) versus three-dimensional (3D) VR environments remains underexplored. Current research primarily utilizes power spectral density (PSD) features to differentiate between 2D and 3D VR conditions, but the potential of other feature parameters for enhanced discrimination is unclear. Additionally, the use of machine learning techniques to classify EEG signals from 2D and 3D VR using alternative features has not been thoroughly investigated, highlighting the need for further research to identify robust EEG features and effective classification methods. Methods: This study recorded EEG signals from participants exposed to 2D and 3D VR video stimuli to investigate the neural differences between these conditions. Key features extracted from the EEG data included PSD and common spatial patterns (CSPs), which capture frequency-domain and spatial-domain information, respectively. To evaluate classification performance, several classical machine learning algorithms were employed: ssupport vector machine (SVM), k-nearest neighbors (KNN), random forest (RF), naive Bayes, decision Tree, AdaBoost, and a voting classifier. The study systematically compared the classification performance of PSD and CSP features across these algorithms, providing a comprehensive analysis of their effectiveness in distinguishing EEG signals in response to 2D and 3D VR stimuli. Results: The study demonstrated that machine learning algorithms can effectively classify EEG signals recorded during watching 2D and 3D VR videos. CSP features outperformed PSD in classification accuracy, indicating their superior ability to capture EEG signals differences between the VR conditions. Among the machine learning algorithms, the Random Forest classifier achieved the highest accuracy at 95.02%, followed by KNN with 93.16% and SVM with 91.39%. The combination of CSP features with RF, KNN, and SVM consistently showed superior performance compared to other feature-algorithm combinations, underscoring the effectiveness of CSP and these algorithms in distinguishing EEG responses to different VR experiences. Conclusions: This study demonstrates that EEG signals recorded during watching 2D and 3D VR videos can be effectively classified using machine learning algorithms with extracted feature parameters. The findings highlight the superiority of CSP features over PSD in distinguishing EEG signals under different VR conditions, emphasizing CSP’s value in VR-induced EEG analysis. These results expand the application of feature-based machine learning methods in EEG studies and provide a foundation for future research into the brain cortical activity of VR experiences, supporting the broader use of machine learning in EEG-based analyses. Full article

Timely and accurate land use information in open-pit mines is essential for environmental monitoring, ecological restoration planning, and promoting sustainable progress in mining regions. This study used high-resolution unmanned aerial vehicle (UAV) imagery, combined with object-oriented methods, optimal segmentation algorithms, and machine learning algorithms, to develop an efficient and practical method for classifying land use in open-pit mines. First, six land use categories were identified: stope, restoration area, building, vegetation area, arterial road, and waters. To achieve optimal scale segmentation, an image segmentation quality evaluation index is developed, emphasizing both high intra-object homogeneity and high inter-object heterogeneity. Second, spectral, index, texture, and spatial features are identified through out-of-bag (OOB) error of random forest and recursive feature elimination (RFE) to create an optimal multi-feature fusion combination. Finally, the classification of open-pit mines was executed by leveraging the optimal feature combination, employing the random forest (RF), support vector machine (SVM), and k-nearest neighbor (KNN) classifiers in a comparative analysis. The experimental results indicated that classification of appropriate scale image segmentation can extract more accurate land use information. Feature selection effectively reduces model redundancy and improves classification accuracy, with spectral features having the most significant effect. The RF algorithm outperformed SVM and KNN, demonstrating superior handling of high-dimensional feature combinations. It achieves the highest overall accuracy (OA) of 90.77%, with the lowest misclassification and omission errors and the highest classification accuracy. The disaggregated data facilitate effective monitoring of ecological changes in open-pit mining areas, support the development of mining plans, and help predict the quality and heterogeneity of raw clay in some areas. Full article

Texture features have been consistently overlooked in digital soil mapping, especially in soil salinization mapping. This study aims to clarify how to leverage texture information for monitoring soil salinization through remote sensing techniques. We propose a novel method for estimating soil salinity content (SSC) that combines spectral and texture information from unmanned aerial vehicle (UAV) images. Reflectance, spectral index, and one-dimensional (OD) texture features were extracted from UAV images. Building on the one-dimensional texture features, we constructed two-dimensional (TD) and three-dimensional (THD) texture indices. The technique of Recursive Feature Elimination (RFE) was used for feature selection. Models for soil salinity estimation were built using three distinct methodologies: Random Forest (RF), Partial Least Squares Regression (PLSR), and Convolutional Neural Network (CNN). Spatial distribution maps of soil salinity were then generated for each model. The effectiveness of the proposed method is confirmed through the utilization of 240 surface soil samples gathered from an arid region in northwest China, specifically in Xinjiang, characterized by sparse vegetation. Among all texture indices, TDT_eI₁ has the highest correlation with SSC (|r| = 0.86). After adding multidimensional texture information, the R² of the RF model increased from 0.76 to 0.90, with an improvement of 18%. Among the three models, the RF model outperforms PLSR and CNN. The RF model, which combines spectral and texture information (SOTT), achieves an R² of 0.90, RMSE of 5.13 g kg⁻¹, and RPD of 3.12. Texture information contributes 44.8% to the soil salinity prediction, with the contributions of TD and THD texture indices of 19.3% and 20.2%, respectively. This study confirms the great potential of introducing texture information for monitoring soil salinity in arid and semi-arid regions. Full article

Mangrove ecosystems play an irreplaceable role in coastal environments by providing essential ecosystem services. Diverse mangrove species have different functions due to their morphological and physiological characteristics. A precise spatial distribution map of mangrove species is therefore crucial for biodiversity maintenance and environmental conservation of coastal ecosystems. Traditional satellite data are limited in fine-scale mangrove species classification due to low spatial resolution and less spectral information. This study employed unmanned aerial vehicle (UAV) technology to acquire high-resolution multispectral and hyperspectral mangrove forest imagery in Guangxi, China. We leveraged advanced algorithms, including RFE-RF for feature selection and machine learning models (Adaptive Boosting (AdaBoost), eXtreme Gradient Boosting (XGBoost), Random Forest (RF), and Light Gradient Boosting Machine (LightGBM)), to achieve mangrove species mapping with high classification accuracy. The study assessed the classification performance of these four machine learning models for two types of image data (UAV multispectral and hyperspectral imagery), respectively. The results demonstrated that hyperspectral imagery had superiority over multispectral data by offering enhanced noise reduction and classification performance. Hyperspectral imagery produced mangrove species classification with overall accuracy (OA) higher than 91% across the four machine learning models. LightGBM achieved the highest OA of 97.15% and kappa coefficient (Kappa) of 0.97 based on hyperspectral imagery. Dimensionality reduction and feature extraction techniques were effectively applied to the UAV data, with vegetation indices proving to be particularly valuable for species classification. The present research underscored the effectiveness of UAV hyperspectral images using machine learning models for fine-scale mangrove species classification. This approach has the potential to significantly improve ecological management and conservation strategies, providing a robust framework for monitoring and safeguarding these essential coastal habitats. Full article

In the United States, several land use and land cover (LULC) data sets are available based on satellite data, but these data sets often fail to accurately represent features on the ground. Alternatively, detailed mapping of heterogeneous landscapes for informed decision-making is possible using high spatial resolution orthoimagery from the National Agricultural Imagery Program (NAIP). However, large-area mapping at this resolution remains challenging due to radiometric differences among scenes, landscape heterogeneity, and computational limitations. Various machine learning (ML) techniques have shown promise in improving LULC maps. The primary purposes of this study were to evaluate bagging (Random Forest, RF), boosting (Gradient Boosting Machines [GBM] and extreme gradient boosting [XGB]), and stacking ensemble ML models. We used these techniques on a time series of Sentinel 2A data and NAIP orthoimagery to create a LULC map of a portion of Irion and Tom Green counties in Texas (USA). We created several spectral indices, structural variables, and geometry-based variables, reducing the dimensionality of features generated on Sentinel and NAIP data. We then compared accuracy based on random cross-validation without accounting for spatial autocorrelation and target-oriented cross-validation accounting for spatial structures of the training data set. Comparison of random and target-oriented cross-validation results showed that autocorrelation in the training data offered overestimation ranging from 2% to 3.5%. The XGB-boosted stacking ensemble on-base learners (RF, XGB, and GBM) improved model performance over individual base learners. We show that meta-learners are just as sensitive to overfitting as base models, as these algorithms are not designed to account for spatial information. Finally, we show that the fusion of Sentinel 2A data with NAIP data improves land use/land cover classification using geographic object-based image analysis. Full article

Soils play a central role in ecosystem functioning, and thus, mapped soil property information is indispensable to supporting sustainable land management. Digital Soil Mapping (DSM) provides a framework to spatially estimate soil properties. However, broad-scale DSM remains challenging because of non-purposively sampled soil data, large data volumes for processing extensive soil covariates, and high model complexities due to spatially varying soil–landscape relationships. This study presents a three-dimensional DSM framework for Switzerland, targeting the soil properties of clay content (Clay), organic carbon content (SOC), pH value (pH), and potential cation exchange capacity (CEC_pot). The DSM approach is based on machine learning and a comprehensive exploitation of soil and remote sensing data archives. Quantile Regression Forest was applied to link the soil sample data from a national soil data base with covariates derived from a LiDAR-based elevation model, from climate raster data, and from multispectral raster time series based on satellite imagery. The covariate set comprises spatially multiscale terrain attributes, climate patterns and their temporal variation, temporarily multiscale land use features, and spectral bare soil signatures. Soil data and predictions were evaluated with respect to different landcovers and depth intervals. All reference soil data sets were found to be spatially clustered towards croplands, showing an increasing sample density from lower to upper depth intervals. According to the R² value derived from independent data, the overall model accuracy amounts to 0.69 for Clay, 0.64 for SOC, 0.76 for pH, and 0.72 for CEC_pot. Reduced model accuracies were found to be accompanied by soil data sets showing limited sample sizes (e.g., CEC_pot), uneven statistical distributions (e.g., SOC), and low spatial sample densities (e.g., woodland subsoils). Multiscale terrain covariates were highly influential for all models; climate covariates were particularly important for the Clay model; multiscale land use covariates showed enhanced importance for modeling pH; and bare soil reflectance was a major driver in the SOC and CEC_pot models. Full article

The identification, classification and mapping of different plant communities and habitats is of fundamental importance for defining biodiversity monitoring and conservation strategies. Today, the availability of high temporal, spatial and spectral data from remote sensing platforms provides dense time series over different spectral bands. In the case of supervised mapping, time series based on classical vegetation indices (e.g., NDVI, GNDVI, …) are usually input characteristics, but the selection of the best index or set of indices (which guarantees the best performance) is still based on human experience and is also influenced by the study area. In this work, several different time series, based on Sentinel-2 images, were created exploring new combinations of bands that extend the classic basic formulas as the normalized difference index. Multivariate Functional Principal Component Analysis (MFPCA) was used to contemporarily decompose the multiple time series. The principal multivariate seasonal spectral variations identified (MFPCA scores) were classified by using a Random Forest (RF) model. The MFPCA and RF classifications were nested into a forward selection strategy to identify the proper and minimum set of indices’ (dense) time series that produced the most accurate supervised classification of plant communities and habitat. The results we obtained can be summarized as follows: (i) the selection of the best set of time series is specific to the study area and the habitats involved; (ii) well-known and widely used indices such as the NDVI are not selected as the indices with the best performance; instead, time series based on original indices (in terms of formula or combination of bands) or underused indices (such as those derivable with the visible bands) are selected; (iii) MFPCA efficiently reduces the dimensionality of the data (multiple dense time series) providing ecologically interpretable results representing an important tool for habitat modelling outperforming conventional approaches that consider only discrete time series. Full article

Canopy fuels determine the characteristics of the entire complex of forest fuels due to their constant changes triggered by the environment; therefore, the development of appropriate strategies for fire management and fire risk reduction requires an accurate description of canopy forest fuels. This paper presents a method for mapping the spatial distribution of canopy fuel loads (CFLs) in alignment with their natural variability and three-dimensional spatial distribution. The approach leverages an object-based machine learning framework with UAV multispectral data and photogrammetric point clouds. The proposed method was developed in the mixed forest of the natural protected area of “Sierra de Quila”, Jalisco, Mexico. Structural variables derived from photogrammetric point clouds, along with spectral information, were used in an object-based Random Forest model to accurately estimate CFLs, yielding R² = 0.75, RMSE = 1.78 Mg, and an average ${Bias}_{rel}$ = 18.62%. Canopy volume was the most significant explanatory variable, achieving a mean decrease in impurity values greater than 80%, while the combination of texture and vegetation indices presented importance values close to 20%. Our modelling approach enables the accurate estimation of CFLs, accounting for the ecological context that governs their dynamics and spatial variability. The high precision achieved, at a relatively low cost, encourages constant updating of forest fuels maps to enable researchers and forest managers to streamline decision making on fuel and forest fire management. Full article

Automatic, efficient, and accurate individual tree species identification and crown parameters extraction is of great significance for biodiversity conservation and ecosystem function assessment. UAV multispectral data have the advantage of low cost and easy access, and hyperspectral data can finely characterize spatial and spectral features. As such, they have attracted extensive attention in the field of forest resource investigation, but their applicability for end-to-end individual tree species identification is unclear. Based on the Mask R-CNN instance segmentation model, this study utilized UAV hyperspectral images to generate spectral thinning data, spectral dimensionality reduction data, and simulated multispectral data, thereby evaluating the importance of high-resolution spectral information, the effectiveness of PCA dimensionality reduction processing of hyperspectral data, and the feasibility of multispectral data for individual tree identification. The results showed that the individual tree species identification accuracy of spectral thinning data was positively correlated with the number of bands, and full-band hyperspectral data were better than other hyperspectral thinning data and PCA dimensionality reduction data, with Precision, Recall, and F1-score of 0.785, 0.825, and 0.802, respectively. The simulated multispectral data are also effective in identifying individual tree species, among which the best result is realized through the combination of Green, Red, and NIR bands, with Precision, Recall, and F1-score of 0.797, 0.836, and 0.814, respectively. Furthermore, by using Green–Red–NIR data as input, the tree crown area and width are predicted with an RMSE of 3.16m² and 0.51m, respectively, along with an rRMSE of 0.26 and 0.12. This study indicates that the Mask R-CNN model with UAV optical images is a novel solution for identifying individual tree species and extracting crown parameters, which can provide practical technical support for sustainable forest management and ecological diversity monitoring. Full article

Progress toward habitat protection goals can effectively be performed using satellite imagery and machine-learning (ML) models at various spatial and temporal scales. In this regard, habitat types and landscape structures can be discriminated against using remote-sensing (RS) datasets. However, most existing research in three-dimensional (3D) habitat mapping primarily relies on same/cross-sensor features like features derived from multibeam Light Detection And Ranging (LiDAR), hydrographic LiDAR, and aerial images, often overlooking the potential benefits of considering multi-sensor data integration. To address this gap, this study introduced a novel approach to creating 3D habitat maps by using high-resolution multispectral images and a LiDAR-derived Digital Surface Model (DSM) coupled with an object-based Random Forest (RF) algorithm. LiDAR-derived products were also used to improve the accuracy of the habitat classification, especially for the habitat classes with similar spectral characteristics but different heights. Two study areas in the United Kingdom (UK) were chosen to explore the accuracy of the developed models. The overall accuracies for the two mentioned study areas were high (91% and 82%), which is indicative of the high potential of the developed RS method for 3D habitat mapping. Overall, it was observed that a combination of high-resolution multispectral imagery and LiDAR data could help the separation of different habitat types and provide reliable 3D information. Full article

The food production system is vulnerable to diseases more than ever, and the threat is increasing in an era of climate change that creates more favorable conditions for emerging diseases. Fortunately, scientists and engineers are making great strides to introduce farming innovations to tackle the challenge. Unmanned aerial vehicle (UAV) remote sensing is among the innovations and thus is widely applied for crop health monitoring and phenotyping. This study demonstrated the versatility of aerial remote sensing in diagnosing yellow rust infection in spring wheats in a timely manner and determining an intervenable period to prevent yield loss. A small UAV equipped with an aerial multispectral sensor periodically flew over, and collected remotely sensed images of, an experimental field in Chacabuco (−34.64; −60.46), Argentina during the 2021 growing season. Post-collection images at the plot level were engaged in a thorough feature-engineering process by handcrafting disease-centric vegetation indices (VIs) from the spectral dimension, and grey-level co-occurrence matrix (GLCM) texture features from the spatial dimension. A machine learning pipeline entailing a support vector machine (SVM), random forest (RF), and multilayer perceptron (MLP) was constructed to identify locations of healthy, mild infection, and severe infection plots in the field. A custom 3-dimensional convolutional neural network (3D-CNN) relying on the feature learning mechanism was an alternative prediction method. The study found red-edge (690–740 nm) and near infrared (NIR) (740–1000 nm) as vital spectral bands for distinguishing healthy and severely infected wheats. The carotenoid reflectance index 2 (CRI2), soil-adjusted vegetation index 2 (SAVI2), and GLCM contrast texture at an optimal distance d = 5 and angular direction θ = 135° were the most correlated features. The 3D-CNN-based wheat disease monitoring performed at 60% detection accuracy as early as 40 days after sowing (DAS), when crops were tillering, increasing to 71% and 77% at the later booting and flowering stages (100–120 DAS), and reaching a peak accuracy of 79% for the spectral-spatio-temporal fused data model. The success of early disease diagnosis from low-cost multispectral UAVs not only shed new light on crop breeding and pathology but also aided crop growers by informing them of a prevention period that could potentially preserve 3–7% of the yield at the confidence level of 95%. Full article

The combination of multi-temporal images and deep learning is an efficient way to obtain accurate crop distributions and so has drawn increasing attention. However, few studies have compared deep learning models with different architectures, so it remains unclear how a deep learning model should be selected for multi-temporal crop classification, and the best possible accuracy is. To address this issue, the present work compares and analyzes a crop classification application based on deep learning models and different time-series data to exploit the possibility of improving crop classification accuracy. Using Multi-temporal Sentinel-2 images as source data, time-series classification datasets are constructed based on vegetation indexes (VIs) and spectral stacking, respectively, following which we compare and evaluate the crop classification application based on time-series datasets and five deep learning architectures: (1) one-dimensional convolutional neural networks (1D-CNNs), (2) long short-term memory (LSTM), (3) two-dimensional-CNNs (2D-CNNs), (4) three-dimensional-CNNs (3D-CNNs), and (5) two-dimensional convolutional LSTM (ConvLSTM2D). The results show that the accuracy of both 1D-CNN (92.5%) and LSTM (93.25%) is higher than that of random forest (~ 91%) when using a single temporal feature as input. The 2D-CNN model integrates temporal and spatial information and is slightly more accurate (94.76%), but fails to fully utilize its multi-spectral features. The accuracy of 1D-CNN and LSTM models integrated with temporal and multi-spectral features is 96.94% and 96.84%, respectively. However, neither model can extract spatial information. The accuracy of 3D-CNN and ConvLSTM2D models is 97.43% and 97.25%, respectively. The experimental results show limited accuracy for crop classification based on single temporal features, whereas the combination of temporal features with multi-spectral or spatial information significantly improves classification accuracy. The 3D-CNN and ConvLSTM2D models are thus the best deep learning architectures for multi-temporal crop classification. However, the ConvLSTM architecture combining recurrent neural networks and CNNs should be further developed for multi-temporal image crop classification. Full article

Imaging spectroscopy is a burgeoning tool for understanding ecosystem functioning on large spatial scales, yet the application of this technology to assess intra-specific trait variation across environmental gradients has been poorly tested. Selection of specific genotypes via environmental filtering plays an important role in driving trait variation and thus functional diversity across space and time, but the relative contributions of intra-specific trait variation and species turnover are still unclear. To address this issue, we quantified the variation in reflectance spectra within and between six uniform stands of Metrosideros polymorpha across elevation and soil substrate age gradients on Hawai‘i Island. Airborne imaging spectroscopy and light detection and ranging (LiDAR) data were merged to capture and isolate sunlit portions of canopies at the six M. polymorpha-dominated sites. Both intra-site and inter-site spectral variations were quantified using several analyses. A support vector machine (SVM) model revealed that each site was spectrally distinct, while Euclidean distances between site centroids in principal components (PC) space indicated that elevation and soil substrate age drive the separation of canopy spectra between sites. Coefficients of variation among spectra, as well as the intrinsic spectral dimensionality of the data, demonstrated the hierarchical effect of soil substrate age, followed by elevation, in determining intra-site variation. Assessments based on leaf trait data estimated from canopy reflectance resulted in similar patterns of separation among sites in the PC space and distinction among sites in the SVM model. Using a highly polymorphic species, we demonstrated that canopy reflectance follows known ecological principles of community turnover and thus how spectral remote sensing addresses forest community assembly on large spatial scales. Full article

Hyperspectral images taken from aircraft or satellites contain information from hundreds of spectral bands, within which lie latent lower-dimensional structures that can be exploited for classifying vegetation and other materials. A disadvantage of working with hyperspectral images is that, due to an inherent trade-off between spectral and spatial resolution, they have a relatively coarse spatial scale, meaning that single pixels may correspond to spatial regions containing multiple materials. This article introduces the Diffusion and Volume maximization-based Image Clustering (D-VIC) algorithm for unsupervised material clustering to address this problem. By directly incorporating pixel purity into its labeling procedure, D-VIC gives greater weight to pixels corresponding to a spatial region containing just a single material. D-VIC is shown to outperform comparable state-of-the-art methods in extensive experiments on a range of hyperspectral images, including land-use maps and highly mixed forest health surveys (in the context of ash dieback disease), implying that it is well-equipped for unsupervised material clustering of spectrally-mixed hyperspectral datasets. Full article