Experimental comparison of support vector machines with random forests for hyperspectral image land cover classification

Considering the form diversity of tree species composition in the Bagong Mountain National Forest Park of China, we mapped tree species utilizing Machine Learning Algorithms (support vector machines (SVM) and random forest (RF) classifiers) based on the OHS-2 hyperspectral satellite image by different datasets which combined spectral information and hyperspectral-derived vegetation indices (VIs) for improving tree species classification and explored the best performance of them. To verify the improvement, the results of physically-based spectral classifiers (spectral angle mapper (SAM) and maximum likelihood (ML) classifiers) were applied to compare with the results of machine learning algorithms. The results indicated an overall accuracy of 94.01%, 96.08%, 82.9% and 79.3% for SVM, RF, SAM and ML classifiers of the best performance using different datasets. Highest accuracies resulted from two machine learning algorithms classifiers; SVM and RF compared to SAM and ML classifiers. Although SVM outperformed RF when using all hyperspectral bands and VIs, the overall accuracy of the RF classifier is higher when compared to the SVM classifier using VIs combined selected features. Meanwhile, the RF classifier performed better than SVM after removing the redundancy of spectral data in training samples. Moreover, the machine learning algorithms successfully classified a small number of tree species (Cedrus deodara and Pterocarya stenoptera C. DC.) in the study area, but the physical spectroscopy-based method failed to classify these species. Such integration strategy improved the effectiveness of enhancing the accuracy of tree species classification and mapping their distribution on broad spatial and temporal scales using machine learning algorithms and hyperspectral imagery.

Land-cover classification using remote sensing imagery is an important part of environmental research because it provides baseline information for ecological vulnerability and risk assessment, disaster management, landscape conservation, local and regional planning, and so on. Rural-land-cover classification is challenging for both object-based image analysis methods and classifiers. The objective of this study is to improve the object-oriented classification accuracy of rural land cover by combining two models based on high spatial resolution imagery. We apply the C5.0 algorithm in R to combine support vector machines (SVMs) and random forest (RF) to create the model RS_C5.0. The effectiveness of the model combination is assessed by comparing the classification results with the state-of-the-art machine learning algorithm, namely extreme gradient boosting (XGBoost). The comparisons are done based on the classification results of both the study area and the case area. Results show that in the classification of the study area, RF performs slightly better than SVM, and XGBoost performs worse than RF but better than SVM. However, in the classification of the case area, SVM performs slightly better than RF and both SVM and RF perform better than XGBoost. Furthermore, RS_C5.0 obtains the highest overall accuracies and kappa coefficients in the classifications of both the study area and the case area. In terms of training time, XGBoost runs the slowest in the classifications of both the study area and the case area. SVM and RF as well as the combined model (RS_C5.0) run much faster than XGBoost classifier. To summarize, the combination of SVM and RF classifiers using C5.0 algorithm is found to be a fast and effective way to improve rural-land-cover classification.

Powdery mildew (PM, Blumeria graminis f. sp. tritici) is a devastating disease for wheat growth and production. It is highly meaningful that the disease severities can be objectively and accurately identified by image visualization technology. In this study, an integral method was proposed based on a hyperspectral imaging dataset and machine learning algorithms. The disease severities of wheat leaves infected with PM were quantitatively identified based on hyperspectral images and image segmentation techniques. A technical procedure was proposed to perform the identification and evaluation of leaf-scale wheat PM, specifically including three primary steps of the acquisition and preprocessing of hyperspectral images, the selection of characteristic bands, and model construction. Firstly, three-dimensional reduction algorithms, namely principal component analysis (PCA), random forest (RF), and the successive projections algorithm (SPA), were comparatively used to select the bands that were most sensitive to PM. Then, three diagnosis models were constructed by a support vector machine (SVM), RF, and a probabilistic neural network (PNN). Finally, the best model was selected by comparing the overall accuracies. The results show that the SVM model constructed by PCA dimensionality reduction had the best result, and the classification accuracy reached 93.33% by a cross-validation method. There was an obvious improvement of the identification accuracy with the model, which achieved an 88.00% accuracy derived from the original hyperspectral images. This study can provide a reference for accurately estimating the disease severity of leaf-scale wheat PM and other plant diseases by non-contact measurement technology.

Land Use and Land Cover (LULC) classification plays a pivotal role in understanding and managing environmental resources. This study presents a novel methodology utilizing sentinel satellite data in conjunction with two robust machine learning algorithms: Random Forest (RF) and Support Vector machine (SVM) on Google Earth Engine platform. Sentinel data, renowned for its high- resolution multispectral imagery, provides rich information for classification. Google Earth Engine provides access to vast geospatial datasets and computational resource, enabling effective analysis. RF and SVM, distinguished for their ability to handle complex dataset, are employed to optimize classification accuracy. A systematic work flow for preprocessing of Sentinel imagery is outlined, followed by implementation of RF and SVM algorithm with a focus on accurately classifying vegetation, built-up areas, barren land and water bodies. Evaluation metrics including overall accuracy and kappa coefficient demonstrate the efficacy of the proposed methodology. A compelling study demonstrates the utility of RF and SVM within GEE for precise LULC mapping, highlighting their pivotal role in supporting informed decision-making for environmental planning and conservation initiatives. Keywords GEE, SVM, RF, Sentinel Data, LULC

Hyperspectral imaging is concerned with the measurement, analysis, and interpretation of spectra acquired from a given scene (or specific object) at a short, medium or long distance by an airbone or satellite sensor. Over the last few years, hyperspectral image data sets have been collected for a great amount of locations over the world, using a variety of instruments for Earth observation. Despite the increasing importance of hyperspectral images in remote sensing applications, there is no common repository of hyperspectral data intended to distribute and share hyperspectral data sets in the community. Quite opposite, the hyperspectral data sets which are available for public use are spread among different storage locations and present significant heterogeneity regarding the storage format, associated meta-data (if any), or ground-truth availability. As a result, the development of a standardized hyperspectral data repository is a highly desired goal in the remote sensing community. In this paper, we take a necessary first step towards the development of a digital repository for remotely sensed hyperspectral data. The proposed system allows uploading new hyperspectral data sets along with meta-data, ground-truth and analysis results, with the ultimate goal of sharing publicly available hyperspectral images within the remote sensing community. The database has been designed in order to allow storing relevant information for the hyperspectral data available through the system, including basic image characteristics (width, height, number of bands, format) and more advanced meta-data (ground-truth information, publications in which the data has been used). The current implementation consists of a front-end to ease the management of images through a web interface, thus containing both synthetic and real hyperspectral images from two highly representative instruments, such as NASAs Airborne Visible Infra-Red Imaging Spectrometer (AVIRIS) over the Cuprite Mining District in Nevada. Most importantly, the developed system includes a spectral unmixing-based content based image retrieval (CBIR) functionality which allows searching for images on the spectral unmixing information (spectrally pure components or endmembers and their associated abundances in the scene). This information is stored as meta-data associated to each hyperspectral image instance, and then used to search and retrieve images based on information content. This paper presents the design of the system and a preliminary validation of the unmixing-based retrieval functionality using both synthetic and real hyperspectral images stored in the database.

In this paper, a contextual Support Vector Machine (SVM) technique based on the principle of Hilbert Space Embedding (HSE) of a local hyperspectral data distribution into an Reproducing Kernel Hilbert Space (RKHS) is proposed to optimally exploit the spectral and local spatial information of the hyperspectral image. The idea of embedding is to map hyperspectral pixels in a local neighborhood into a single point in the RKHS that can uniquely represent those pixels collectively. Previously, the authors have employed an HSE called empirical mean map to build the contextual SVM. In this work, a weighted empirical mean map is utilized to exploit the similarities and variation in the local spatial information. For every pixel, a small set of the neighboring pixels in a hyperspectral image are mapped into an RKHS induced by a certain kernel (Eg. Gaussian RBF kernel) and then, the embedded point of these group of pixels is obtained by calculating the weighted empirical mean of these mapped points. The weights are determined based on the distance between the pixel in consideration and its neighbors. An SVM separating hyperplane is built to maximize the margin between classes formed by weighted empirical means. The proposed technique showed significant improvement over the existing contextual and composite kernels on two hyperspectral image data sets.

We discriminated different successional forest stages, forest degradation, and land use classes in the Tapajós National Forest (TNF), located in the Central Brazilian Amazon. We used full polarimetric images from ALOS/PALSAR-2 that have not yet been tested for land use and land cover (LULC) classification, neither for forest degradation classification in the TNF. Our specific objectives were: (1) to test the potential of ALOS/PALSAR-2 full polarimetric images to discriminate LULC classes and forest degradation; (2) to determine the optimum subset of attributes to be used in LULC classification and forest degradation studies; and (3) to evaluate the performance of Random Forest (RF) and Support Vector Machine (SVM) supervised classifications to discriminate LULC classes and forest degradation. PALSAR-2 images from 2015 and 2016 were processed to generate Radar Vegetation Index, Canopy Structure Index, Volume Scattering Index, Biomass Index, and Cloude–Pottier, van Zyl, Freeman–Durden, and Yamaguchi polarimetric decompositions. To determine the optimum subset, we used principal component analysis in order to select the best attributes to discriminate the LULC classes and forest degradation, which were classified by RF. Based on the variable importance score, we selected the four first attributes for 2015, alpha, anisotropy, volumetric scattering, and double-bounce, and for 2016, entropy, anisotropy, surface scattering, and biomass index, subsequently classified by SVM. Individual backscattering indexes and polarimetric decompositions were also considered in both RF and SVM classifiers. Yamaguchi decomposition performed by RF presented the best results, with an overall accuracy (OA) of 76.9% and 83.3%, and Kappa index of 0.70 and 0.80 for 2015 and 2016, respectively. The optimum subset classified by RF showed an OA of 75.4% and 79.9%, and Kappa index of 0.68 and 0.76 for 2015 and 2016, respectively. RF exhibited superior performance in relation to SVM in both years. Polarimetric attributes exhibited an adequate capability to discriminate forest degradation and classes of different ecological succession from the ones with less vegetation cover.

Spectral-spatial classification of hyperspectral data with mutual information based segmented stacked autoencoder approach

Optical image provides a unique spectral and textural opportunity to acquire data free of charge in cloud-free weather conditions. However, textural information from recent Sentinel-2A data has not been sufficiently recognized. Several researches have shown that texture features are essential for extracting crop information. This study presents a new scheme that integrates spectral bands and textural features for mapping maize fields and other land cover types (i.e., build-up, trees, others, and water) and evaluates the effectiveness of random forest (RF) and support vector machine (SVM) classifiers with different kernel functions based on Sentinel 2A (S-2A) data. The textural features were derived from multiple bands of S-2A image using a grey-level co-occurrence matrix (GLCM) algorithm. To normalize the analysis of these features, Random Forest feature selection was employed. The outcomes show that S-2A images are appropriate for mapping maize fields and other land covers when the image is accessible during the growing season. The SVM classifier with a radial basis function (RBF) kernel is superior to a polynomial (P), linear (L), and sigmoid (S) kernel functions, and the RF classifiers. For the SVM_RBF, the highest overall accuracy was obtained from the integration of spectral bands and textural features at scenario 2 SVM_RBF (87.40%), followed by the combination of spectral and textural features at scenario 2 SVM_P kernel (86.56%). However, scenario 5 SVM_P kernel (85.54%), scenario 4 SVM_P kernel function (85.30%), the spectral features alone, and the SVM_P kernel function, have the highest O/A (83.90%) than that of the SVM_RBF (83.66%). The inclusion of textural features contributed significantly in mapping the smallholder’s maize field in the study area. This demonstrated the value of S-2A in agricultural research.

This paper proposes a new distributed architecture for supervised classification of large volumes of earth observation data on a cloud computing environment. The architecture supports distributed execution, network communication, and fault tolerance in a transparent way to the user. The architecture is composed of three abstraction layers, which support the definition and implementation of applications by researchers from different scientific investigation fields. The implementation of architecture is also discussed. A software prototype (available online), which runs machine learning routines implemented on the cloud using the Waikato Environment for Knowledge Analysis (WEKA), a popular free software licensed under the GNU General Public License, is used for validation. Performance issues are addressed through an experimental analysis in which two supervised classifiers available in WEKA were used: random forest and support vector machines. This paper further describes how to include other classification methods in the available software prototype.

Researchers constantly seek more efficient detection techniques to better utilize enhanced image resolution in accurately detecting and monitoring land cover. Recently, convolutional neural networks (CNNs) have shown high performances comparable to or even better than widely used and adopted machine learning techniques. The aim of this study is to investigate the application of CNNs for land cover classification by using two-dimensional (2-D) spectral curve graphs from multispectral satellite images. The land cover classification was conducted in Concord, New Hampshire, USA, and South Korea by using multispectral images acquired from 30-m Landsat-8 and 500-m Geostationary Ocean Color Imager images. For the construction of input data specific to CNNs, two seasons (winter and summer) of multispectral bands were transformed into 2-D spectral curve graphs for each class. Land cover classification results of CNNs were compared with the results of support vector machines (SVMs) and random forest (RFs). The CNNs model showed higher performance than RFs and SVMs in both study sites. The examination of land cover classification maps demonstrates a good agreement with reference maps, Google Earth images, and existing global scale land cover map, especially for croplands. Using the spectral curve graph could incorporate the phenological cycles on classifying the land cover types. This study shows that the use of a new transformation of spectral bands into a 2-D form for application in CNNs can improve land cover classification performance.

This study proposes a workflow for land use and land cover (LULC) classification of Advanced Land Observing Satellite-2 (ALOS-2) Phased Array type L-band Synthetic Aperture Radar-2 (PALSAR-2) images of the Brazilian tropical savanna (Cerrado) biome. The following LULC classes were considered: forestlands; shrublands; grasslands; reforestations; croplands; pasturelands; bare soils/straws; urban areas; and water reservoirs. The proposed approach combines polarimetric attributes, image segmentation, and machine-learning procedures. A set of 125 attributes was generated using polarimetric ALOS-2/PALSAR-2 images, including the van Zyl, Freeman–Durden, Yamaguchi, and Cloude–Pottier target decomposition components, incoherent polarimetric parameters (biomass indices and polarization ratios), and HH-, HV-, VH-, and VV-polarized amplitude images. These attributes were classified using the Naive Bayes (NB), DT J48 (DT = decision tree), Random Forest (RF), Multilayer Perceptron (MLP), and Support Vector Machine (SVM) algorithms. The RF, MLP, and SVM classifiers presented the most accurate performances. NB and DT J48 classifiers showed a lower performance in relation to the RF, MLP, and SVM. The DT J48 classifier was the most suitable algorithm for discriminating urban areas and natural vegetation cover. The proposed workflow can be replicated for other SAR images with different acquisition modes or for other types of vegetation domains.

Accurate Land Use and Land Cover (LU/LC) classification is essential for sustainable resource management, urban development, and environmental conservation. The integration of remote sensing data with supervised machine learning algorithms has significantly enhanced classification accuracy and efficiency. This study evaluates the performance of five widely used supervised learning algorithms namely 1) Classification and Regression Tree (CART), 2) Gradient Boost Tree (GB), 3) K-Nearest Neighbours (KNN), 4) Support Vector Machine (SVM) and 5) Random Forest (RF) for LU/LC mapping in study area of East Godavari District, Andhra Pradesh, India over a time period of 2 years between 2023 and 2025. High-resolution Landsat-8 imagery is processed and classified using above algorithms, with model performance assessed based on overall accuracy, Kappa coefficient, precision and F1-score. The findings indicated that Gradient Tree Boost demonstrated superior performance compared to the other classifiers, attaining the highest accuracy of 98.26% along with a Kappa coefficient of 0.9761. Random Forest closely followed, achieving an accuracy of 97.39% and a Kappa value of 0.9642. Additionally, both SVM and KNN exhibited strong classification capabilities, with respective accuracies of 96.52% and Kappa values of 0.9522, highlighting their effectiveness in land cover classification applications. The study also examines the computational efficiency and reliability of each classifier, offering insights into their suitability for LU/LC analysis in diverse landscapes. The findings contribute to the optimization of machine learning techniques for remote sensing applications, aiding in data-driven decision-making for land management. Future research can explore deep learning-based classification models and multi-temporal analysis to further enhance LU/LC mapping accuracy.

Deep learning techniques have boosted the performance of hyperspectral image (HSI) classification. In particular, convolutional neural networks (CNNs) have shown superior performance to that of the conventional machine learning algorithms. Recently, a novel type of neural networks called capsule networks (CapsNets) was presented to improve the most advanced CNNs. In this paper, we present a modified two-layer CapsNet with limited training samples for HSI classification, which is inspired by the comparability and simplicity of the shallower deep learning models. The presented CapsNet is trained using two real HSI datasets, i.e., the PaviaU (PU) and SalinasA datasets, representing complex and simple datasets, respectively, and which are used to investigate the robustness or representation of every model or classifier. In addition, a comparable paradigm of network architecture design has been proposed for the comparison of CNN and CapsNet. Experiments demonstrate that CapsNet shows better accuracy and convergence behavior for the complex data than the state-of-the-art CNN. For CapsNet using the PU dataset, the Kappa coefficient, overall accuracy, and average accuracy are 0.9456, 95.90%, and 96.27%, respectively, compared to the corresponding values yielded by CNN of 0.9345, 95.11%, and 95.63%. Moreover, we observed that CapsNet has much higher confidence for the predicted probabilities. Subsequently, this finding was analyzed and discussed with probability maps and uncertainty analysis. In terms of the existing literature, CapsNet provides promising results and explicit merits in comparison with CNN and two baseline classifiers, i.e., random forests (RFs) and support vector machines (SVMs).

Deep learning based methods have recently been successfully explored in hyperspectral image classification field. However, training a deep learning model still requires a large number of labeled samples, which is usually impractical in hyperspectral images. In this paper, a simple but effective feature extraction method is proposed for hyperspectral image classification. Specifically, a pretrained deep convolutional neural network based on the ImageNet dataset is used to extract spatial features of a hyperspectral image. Recently, it is easy to obtain a pretrained convolutional neural network on the Internet. Note that the pretrained models are trained by using the ImageNet dataset. This means that the proposed method does not need labeled hyperspectral samples to train the deep model. Therefore, the proposed method alleviates the problem of lacking labeled samples and avoids the artificial design of feature extraction rules. Finally, the extracted features are stacked with spectral features as the input of a support vector machine classifier. The proposed method is conducted on three widely used hyperspectral image datasets. The experimental results demonstrate that the proposed method could outperform the conventional feature extraction methods and deep learning based methods.