An Overview on Remote Sensing Image Classification Methods with a Focus on Support Vector Machine

AN ASSESSMENT OF SPATIAL VARIATION OF LAND SURFACE CHARACTERISTICS OF MINNA, NIGER STATE NIGERIA FOR SUSTAINABLE URBANIZATION USING GEOSPATIAL TECHNIQUES

Remote Sensing (RS) image classification has recently attracted great attention for its application in different tasks, including environmental monitoring, battlefield surveillance, and geospatial object detection. The best practices for these tasks often involve transfer learning from pre-trained Convolutional Neural Networks (CNNs). A common approach in the literature is employing CNNs for feature extraction, and subsequently train classifiers exploiting such features. In this paper, we propose the adoption of transfer learning by fine-tuning pre-trained CNNs for end-to-end aerial image classification. Our approach performs feature extraction from the fine-tuned neural networks and remote sensing image classification with a Support Vector Machine (SVM) model with linear and Radial Basis Function (RBF) kernels. To tune the learning rate hyperparameter, we employ a linear decay learning rate scheduler as well as cyclical learning rates. Moreover, in order to mitigate the overfitting problem of pre-trained models, we apply label smoothing regularization. For the fine-tuning and feature extraction process, we adopt the Inception-v3 and Xception inception-based CNNs, as well the residual-based networks ResNet50 and DenseNet121. We present extensive experiments on two real-world remote sensing image datasets: AID and NWPU-RESISC45. The results show that the proposed method exhibits classification accuracy of up to 98%, outperforming other state-of-the-art methods.

Remote sensing classification, as an important means of urban planning and construction, has been widely concerned. Urban land use classification is extremely challenging tasks because of some land covers are spectrally too similar to be separated using only the spectral information of remote sensing image. Object-oriented remote sensing image classification method overcomes the drawbacks of traditional pixel-based classification method. It combines the spectral, special structure and texture features of the images, can effectively avoid the phenomenon of "different objects share the same spectrum" or "the same objects differ in spectrum. Support Vector Machine (SVM) is an excellent tool for remote sensing classification. Combination of both can develop their own advantages to do high-resolution remote sensing image classification. Using a public image in Harbin city as an example, classification based on object-oriented method and SVM has achieved better results than traditional pixel-based classification method.

Abstract: The classification of images can be influenced by various factors. When multiple sources of data are available for earth exploration, extracting valuable information or achieving effective outcomes can be both fascinating and challenging. Remote Sensing Images (RSI) classification plays a crucial role in remote sensing applications and serves as the foundation for efficient classification. Various classifiers are employed to classify different types of targets in remote sensing images. Achieving improved target classification is of utmost importance in both military and civilian domains. This paper focuses on summarizing advanced approaches used to enhance classification accuracy. The findings indicate that the Convolutional Neural Network surpasses all other conventional classification methods.

Accuracy of land cover classification in remotely sensed images relies on the features extracted and the classifier used. Texture features are significant in land cover classification. Traditional texture models capture only patterns with discrete boundaries whereas fuzzy patterns need to be classified by assigning due weightage to uncertainty. When remotely sensed image contains noise, the image may have fuzzy patterns characterizing land covers and fuzzy boundaries separating land covers. So a fuzzy texture model is proposed for effective classification of land covers in remotely sensed images and the model uses Sugeno Fuzzy Inference System (FIS). Support Vector Machine (SVM) is used for precise and fast classification of image pixels. Hence it is proposed to use a hybrid of fuzzy texture model and SVM for land cover classification of remotely sensed images. In this chapter, land cover classification of IRS-P6, LISS-IV remotely sensed image is performed using multivariate version of the proposed texture model.

To overcome the difficulty of automating and intelligently classifying the ground features in remote-sensing hyperspectral images, machine learning methods are gradually introduced into the process of remote-sensing imaging. First, the PaviaU, Botswana, and Cuprite hyperspectral datasets are selected as research subjects in this study, and the objective is to process remote-sensing hyperspectral images via machine learning to realize the automatic and intelligent classification of features. Then, the basic principles of the support vector machine (SVM) and extreme learning machine (ELM) classification algorithms are introduced, and they are applied to the datasets. Next, by adjusting the parameter estimates using a restricted Boltzmann machine (RBM), a new terrain classification model of hyperspectral images that is based on a deep belief network (DBN) is constructed. Next, the SVM, ELM, and DBN classification algorithms for hyperspectral image terrain classification are analysed and compared in terms of accuracy and consistency. The results demonstrate that the average detection accuracies of ELM on the three datasets are 89.54%, 96.14%, and 96.28%, and the Kappa coefficient values are 0.832, 0.963, and 0.924; the average detection accuracies of SVM are 88.90%, 92.11%, and 91.68%, and the Kappa coefficient values are 0.768, 0.913, and 0.944; the average detection accuracies of the DBN classification model are 92.36%, 97.31%, and 98.84%, and the Kappa coefficient values are 0.883, 0.944, and 0.972. The results also demonstrate that the classification accuracy of the DBN algorithm exceeds those of the previous two methods because it fully utilizes the spatial and spectral information of hyperspectral remote-sensing images. In summary, the DBN algorithm that is proposed in this study has high application value in object classification for remote-sensing hyperspectral images.

Abstract. Image processing has been one of the efficient technologies for GIS data requisition. Support Vector Machines (SVMs) have peculiar advantages in handling problems with small sample sizes, nonlinearity, and high dimensionality. However, SVMs can only solve two-class problems while multi-class decision is impossible. Error correcting output coding (ECOC) SVMs enhance the ability of fault tolerance when solving multi-class classification problems, which makes ECOC SVMs suitable for remote sensing image classification. In this paper, the generalization ability of ECOC SVMs is discussed. ECOC SVMs with optimum coding matrices are selected by experiment, and applied to remote sensing image classification. Experimental results show that, compared with Conventional multi-class classification methods, less SVM sub-classifiers are needed for ECOC SVMs in remote sensing image classification, and the classification accuracy is also improved.

It is a challenge to obtain accurate result in remote sensing images classification, which is affected by many factors. In this paper, aiming at correctly identifying land use types reflec ted in remote sensing images, support vector machine, maximum likelihood classifier, backpropagation neural network, fuzzy c-means, and minimum distance classifier were combined to construct three multiple classifier systems (MCSs). Two MCSs were implemented, namely, comparative major voting (CMV) and Bayesian average (BA). One method called WA-AHP was proposed, which introduced analytic hierarchy process into MCS. Classification results of base classifiers and MCSs were compared with the ground truth map. Accuracy indicators were computed and receiver operating characteristic curves were illustrated, so as to evaluate the performance of MCSs. Experimental results show that employing MCSs can increase classification accuracy significantly, compared with base classifiers. From the accuracy evaluation result and visual check, the best MCS is WA-AHP with overall accuracy of 94.2%, which overmatches BA and rivals CMV in this paper. The producer’s accuracy of each land use type proves the good performance of WA-AHP. Therefore, we can draw the conclusion that MCS is superior to base classifiers in remote sensing image classification, and WA-AHP is an efficient MCS.

With the continuous development of remote sensing technology, the data volume of high-resolution is increasing with the large coverage of high-resolution remote sensing images, changeable objects, and complex backgrounds. However, the sensitivity field of current convolutional neural networks is relatively small. This makes it difficult to capture information from the global context. Therefore, we propose remote sensing image classification with a Detailed Attention scheme and a Teacher-Student network named DATS to capture information in the global context. Firstly, the detailed attention scheme is used to integrate the spatial relationship of the feature graph into the feature channel. Thus feature graph is transformed into an attention map, generating structure-preserving and detail-preserving images. Then, the teacher-student network takes detail-preserving and structure-preserving images as inputs and uses feature refiners to enhance the fine-grained details of the images. Finally, fine-grained details learned from the teacher network are integrated into the main network by knowledge distillation, which achieves effective integration both in local detail features and global structure features. Experiments on FGSCR-42, WHU-RS19, and NWPU data sets showed that the Top-1 classification accuracy of our method reached 88.82%, 91.82%, and 87.60%, respectively.

Traditional remote sensing image surveys rely on field observations, which have shortcomings such as low efficiency and limited coverage, making it difficult to meet the needs of urban planning and management. In response to this, this article conducts in-depth research on remote sensing image classification methods based on graph neural networks (GNN). This study adopts a joint framework of GAT (Graph Attention Network)+SLC (Simple Linear Iterative Clustering)+Hybrid Pooling. Use SLIC algorithm to perform superpixel segmentation on hyperspectral data and construct a graph structure that integrates spectral spatial features. Utilizing GAT's attention mechanism to enhance feature interaction among neighboring nodes, combined with a hybrid pooling strategy to balance feature dimensionality reduction and key information preservation. The experiment was conducted on the public datasets WHU-RS19 and UC Merced, using traditional CNN, random forest, and minimum distance methods as baseline models. The significance of the differences was verified through independent sample t-test. By expanding the sample coverage, optimizing the model regularization strategy, and continuously iterating training, the generalization performance of the model has been significantly improved. Therefore, from the comparison of classification performance, graph NN classification is superior to minimum distance classification. This method utilizes perspective remote sensing data to achieve accurate ground target classification and dynamic change information extraction, providing reliable data support for urban land survey and management.

Remote Sensing (RS) image classification, particularly involving Earth Observation (EO) satellite data, presents significant challenges due to the complexity and variety of image content. This study addresses these challenges by evaluating the performance of three advanced deep learning models—DenseNet121, ResNet50, and EfficientNetB7—on the UC Merced Land Use (UCM) dataset. By leveraging pre-trained Convolutional Neural Networks (CNNs) through transfer learning, our approach effectively mitigates the issue of limited labelled data and enhances the accuracy of classification in high-resolution aerial imagery. This paper provides an in-depth study of said models, emphasizing their accuracy, precision, recall, and computing efficiency in the classification of land use domains. The findings provide insightful information about how well these various CNN architectures perform in classifying remote sensing images and lay the groundwork for further deep learning-based land use categorization research

Land cover classification is a complex process that may be affected by many factors. Since the first resource satellite was launched in 1972, the remote sensing community has witnessed the impressive progress in image classification methods, which is primarily driven by the advancement of remote sensing technology and computer technology. In recent years, non-parametric classifiers such as the neural network, the decision tree classifier and other classifiers have developed increasingly. Therefore, the four broadly used classification methods, which are Maximum Likelihood Classification (MLC), Self-Organized Neural Network (SONN), Support Vector Machine (SVM) and Decision Tree Classification (DTC) are firstly applied for the land cover remote sensing classification based on the multitemporal Moderate Resolution Imaging Spectroradiometer (MODIS) images of Heilongjiang area. The emphasis is placed on the comparison of the four classification methods and the techniques used for improving classification accuracy. Then, we compare the four classifiers through different aspects. Through the comparison, we got the conclusions: DTC is the best, and MLC as one of the classical methods is more stable than other three methods. Therefore, we make the land cover classification test over China using DTC and MLC methods and compare them again. We also believe that the conclusions we got in this paper are valuable for how to select an appropriate classifier in the similar applications.

With reference to the literature worldwide, it is obvious that Support Vector Machine (SVM), a machine learning algorithm has proven records for excellent results regarding Classification of Image. But, Remote Sensing Images are considered as most complex in nature as far as classification is concern. Supervised classification of Remote Sensing Images needs more precise machine learning models, which will be fast and efficient. SVM do satisfy researchers all over the world as far as Remote Sensing Images are concern. Basically, SVM is non-parametric statistical learning based model, which acts like binary classifier. SVM represents a group of superior machine learning algorithms, where it decomposes the parameter of the problem into a quadratic optimization technique. Hence, SVM is used to locate optimum boundaries between classes, which in return generalize to unseen samples with least error among all possible boundaries separating two classes. SVM uses density estimation function for developing easy and efficient learning parameters. Like other supervised algorithms, SVM also undergo into Training, Learning and Testing Phase for classifying any image. Besides all parameters, training sample selection and optimization is crucial part that affects the classification accuracy of remote sensing images. We need to address this issue in our project so as to devise noble algorithm or approach, which could make SVM, a more robust statistical learning model.

ABSTRACTLand-cover mapping is an important research topic with broad applicability in the remote-sensing domain. Machine learning algorithms such as Maximum Likelihood Classifier (MLC), Support Vector Machine (SVM), Artificial Neural Network (ANN), and Random Forest (RF) have been playing an important role in this field for many years, although deep neural networks are experiencing a resurgence of interest. In this article, we demonstrate early efforts to apply deep learning-based classification methods to large-scale land-cover mapping. Based on the Stacked Autoencoder (SAE), one of the deep learning models, we built a classification framework for large-scale remote-sensing image processing. We adjusted and optimized the model parameters based on our test samples. We compared the performance of the SAE-based approach with traditional classification algorithms including RF, SVM, and ANN with multiple performance analytics. Results show that the SAE classifier trained with an entire set of African training samples achieves an overall classification accuracy of 78.99% when assessed by test samples collected independently of training samples, which is higher than the accuracies achieved by the other three classifiers (76.03%, 77.74%, and 77.86% of RF, SVM, and ANN, respectively) based on the same set of test samples. We also demonstrated the advantages of SAE in prediction time and land-cover mapping results in this study.

The support vector machine (SVM) has been a dominant machine-learning technique in the last decade and has demonstrated its efficiency in many applications. Research on classification of hyperspectral images have shown the efficiency of this method to overcome the Hughes phenomenon for classification of such images. A major drawback of classification by SVM is that this classifier was originally developed to solve binary problems, and the algorithms for multiclass problems usually have a high-computational load. In this article, a new and fast method for multiclass problems is proposed. This method has two stages. In the first stage, samples are classified by a maximum likelihood (ML) classifier, and in the second stage, SVM selects the final label of a sample among high-probability classes for that sample by a tree structure. So, for each sample, only some classes must be searched by SVM to find its label. The uncertainty of ML classification for a sample is obtained by the entropy of probabilities, and the number of classes that must be searched by SVM for a sample is obtained based on the uncertainty of that sample in the primary ML classification. This approach is compared with two widely used multiclass algorithms: one-against-one (OAO) and directed acyclic graph (DAGSVM). The obtained results on real data from the Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) revealed less computational time and better accuracy compared to these multiclass algorithms.