Decision Boundary Feature Extraction Research Articles

Spectral and spatial features-based HSI classification using multiple neuron-based learning approach

With the improvement of remote sensing application, hyperspectral images are used in many applications. Fusion of spatial and spectral data is an actual way in improving the accuracy of hyper-spectral image classification. In this work, we proposed spectral data with spatial details based on hyper-spectral image classification method using neural network classifiers and multi neurons-based learning approach. This is used to classify the remote sensing images with specific class labels. The features may be supernatural and latitudinal data is extracted using boundary values using decision boundary feature extraction (DBFE); which are trained using convolutional neural networks (CNN) to improve the accuracy in labelling the classes. The methodology entails of training with adding regulariser towards the loss function recycled, train the neural networks.

Spectral and spatial features-based HSI classification using multiple neuron-based learning approach

With the improvement of remote sensing application, hyperspectral images are used in many applications. Fusion of spatial and spectral data is an actual way in improving the accuracy of hyper-spectral image classification. In this work, we proposed spectral data with spatial details based on hyper-spectral image classification method using neural network classifiers and multi neurons-based learning approach. This is used to classify the remote sensing images with specific class labels. The features may be supernatural and latitudinal data is extracted using boundary values using decision boundary feature extraction (DBFE); which are trained using convolutional neural networks (CNN) to improve the accuracy in labelling the classes. The methodology entails of training with adding regulariser towards the loss function recycled, train the neural networks.

Incremental feature extraction based on decision boundaries

Feature extraction is a key algorithm to solve the dimensionality problem. Most feature extraction algorithms use a batch mode, which requires all data available at the same time to calculate new features. Recently, with more available data and advancements of transmission technology, the need for incremental algorithms has increased. In this paper, we propose a gradient descent DBFE method (GDDBFE) that shows a substantial improvement in processing time. Based on this GDDBFE, we then propose an incremental gradient descent decision boundary feature extraction method (IGDDBFE). The proposed IGDDBFE method consists of two steps: updating the decision boundaries and adding discriminately informative features with newly added samples and then updating the feature vectors by incremental eigenvector updates. Experiments with real-world databases show that the proposed method shows improved performance compared to some existing methods.

AN EXTENDED SPECTRAL–SPATIAL CLASSIFICATION APPROACH FOR HYPERSPECTRAL DATA

Abstract. In this paper an extended classification approach for hyperspectral imagery based on both spectral and spatial information is proposed. The spatial information is obtained by an enhanced marker-based minimum spanning forest (MSF) algorithm. Three different methods of dimension reduction are first used to obtain the subspace of hyperspectral data: (1) unsupervised feature extraction methods including principal component analysis (PCA), independent component analysis (ICA), and minimum noise fraction (MNF); (2) supervised feature extraction including decision boundary feature extraction (DBFE), discriminate analysis feature extraction (DAFE), and nonparametric weighted feature extraction (NWFE); (3) genetic algorithm (GA). The spectral features obtained are then fed into the enhanced marker-based MSF classification algorithm. In the enhanced MSF algorithm, the markers are extracted from the classification maps obtained by both SVM and watershed segmentation algorithm. To evaluate the proposed approach, the Pavia University hyperspectral data is tested. Experimental results show that the proposed approach using GA achieves an approximately 8 % overall accuracy higher than the original MSF-based algorithm.

A Comparative Performance Analysis of Feature Extraction Techniques for Hyperspectral Image Classification

Obtaining thematic maps using image classification techniques from hyperspectral datasets is a very difficult image processing task. In hyperspectral image analysis dimensionality reduction is one of the challenging pre-processing tasks, which is achieved using feature extraction techniques. The beauty of these techniques is that they drastically reduces the dimensionality of image and at the same time preserves the majority of the essential information. In this paper few most frequently used dimensionality reduction methods are being investigated, which helps to get accurateness. This research work presents a relative performance investigation of few mostly frequently used feature extraction techniques like Decision Boundary Feature Extraction (DBFE), Non-Parametric Weighted Feature Extraction (NWFE), Discriminative analysis feature extraction (DAFE) and Principal Component Analysis (PCA). The classification is carried out using two most widely used classification techniques including Gaussian Maximum Likelihood (GML) and neural network (NNs). The results obtained after performing experiments indicates that Decision Boundary Feature Extraction (DBFE) technique has provided the best accuracy among various investigated feature extraction techniques. The application areas of this research include areas like identification of exact location in battle field, drought affected areas, flooded areas and weather forecasting etc.

Noise Effects on Feature Mining Non-Parametric Supervised Feature Extraction Techniques

In this paper two famous and commonly used feature mining non-parametric supervised feature extraction techniques (NSFETs) called Non-parametric Weighted Feature Extraction (NWFE) and Decision Boundary Feature Extraction (DBFE) are analyzed to see their efficiency in the presence of noise. In particularly these feature extraction techniques are used in classification as they give better classification accuracy. This study reveals that NSFETs are very sensitive to noise because of which the number of features increases and we get low classification accuracy. In order to see the behavior of NSFETs, spatial and spectral information from hyperspectral image classification is used. The experimental results show that in the presence of noise, spectral information is much more effected than the spatial information when features are extracted using the NSFETs. It is also examined that NWFE is more affected by noise than DBFE. The linear filtering technique is used just before the classifier in order to mitigate the noise effects in NSFETs. Using linear filtering just before the classifier does improve the final classification accuracy but with high number of spatial and spectral features. This does not satisfy the one of the main purpose of feature extraction and that is feature reduction.

Hybrid Model with Super Resolution and Decision Boundary Feature Extraction and Rule based Classification of High Resolution Data

Data Mining is defined as the process of applying specific algorithms for extracting patterns from data. Data mining strategies can be grouped into Classification, Prediction, Estimation, Association Rule Mining and Clustering. Satellite images are used these days. With the help of high resolution image we can extract the features of an image. There are many types of satellites which provide accurate and high resolution images to individuals and organizations. High resolution images give detail information about rural, land, buildings, streets etc. Applications of High resolution images are defense, disaster management and medicals. Object identification is to identify an object which belongs to and feature extraction to extract features from the image like size, shape, texture and spectral information of the image. With the help of feature extraction techniques we can easily find the relevant information from the data and the data can easily be classified. In this hybrid error super resolution model and decision boundary feature extraction algorithms are used and then both the algorithms are compared. This gives the better accuracy with the help of kappa coefficient. General Terms Classification, Feature Extraction, Super Resolution.

Classification of hyperspectral data using extended attribute profiles based on supervised and unsupervised feature extraction techniques

The classification of remote sensing data based on the exploitation of spatial features extracted with morphological and attribute profiles has been recently gaining importance. With the development of efficient algorithms to construct the profiles for large datasets, such methods are becoming even more relevant. When dealing with hyperspectral imagery, the profiles are traditionally built on the first few principal components computed from the data. However, it needs to be determined if other feature reduction approaches are better suited to create base images for the profiles. In this article, we explore the use of profiles based on features derived from three supervised feature extraction techniques (i.e. Discriminant Analysis Feature Extraction, Decision Boundary Feature Extraction and Non-parametric Weighted Feature Extraction) and two unsupervised feature-extraction techniques (i.e. Principal Component Analysis (PCA) and Kernel PCA) in classification and compare the classification accuracies obtained by using different techniques for two different classification methods. The obtained results indicate significant improvements in the accuracies using the supervised feature extraction methods. However, the choice of the method affects the quality of the results for different datasets depending on the availability of the training samples.

On the influence of feature reduction for the classification of hyperspectral images based on the extended morphological profile

In this study we investigated the classification of hyperspectral data with high spatial resolution. Previously, methods that generate a so-called extended morphological profile (EMP) from the principal components of an image have been proposed to create base images for morphological transformations. However, it can be assumed that the feature reduction (FR) may have a significant effect on the accuracy of the classification of the EMP. We therefore investigated the effect of different FR methods on the generation and classification of the EMP of hyperspectral images from urban areas, using a machine learning-based algorithm for classification. The applied FR methods include: principal component analysis (PCA), nonparametric weighted feature extraction (NWFE), decision boundary feature extraction (DBFE), Gaussian kernel PCA (KPCA) and Bhattacharyya distance feature selection (BDFS). Experiments were run with two classification algorithms: the support vector machine (SVM) and random forest (RF) algorithms. We demonstrate that the commonly used PCA approach seems to be nonoptimal in a large number of cases in terms of classification accuracy, and the other FR methods may be more suitable as preprocessing approaches for the EMP.

Object-oriented subspace analysis for airborne hyperspectral remote sensing imagery

An object-oriented mapping approach based on subspace analysis of airborne hyperspectral images was investigated in this paper. Hyperspectral features were extracted based on subspace learning approaches, in order to reduce the redundancy of spectral space and extract the characteristic images for the further object-oriented classification. In this paper, three kinds of spectral feature extraction (FE) methods were utilized to obtain the subspace of airborne hyperspectral data: (1) unsupervised FE, such as PCA (principal component analysis), ICA (independent component analysis) and MNF (maximum noise fraction); (2) supervised FE, e.g. DBFE (decision boundary feature extraction), DAFE (discriminant analysis feature extraction) and NWFE (nonparametric weighted feature extraction); and (3) linear mixture analysis. Afterwards, the extracted subspace features were fed into the object-based classification system. The FNEA (fractal net evolution approach) was utilized to extract objects from the subspace images and SVM (support vector machines) was then used to classify the object-based features. Experiments were conducted on two airborne hyperspectral datasets: (1) the AVIRIS dataset over the northwest Indiana's Pine with 220 spectral bands (agricultural region), and (2) the ROSIS dataset over Pavia University, northern of Italy with 102 spectral bands (urban region). Results revealed that the proposed object-based approach could give significantly higher accuracies than the traditional pixel-based subspace classification.

Estimation of abundance and distribution of two moist tall grasses in the Watarase wetland, Japan, using hyperspectral imagery

The dominant grasses in a wetland are of critical concern for the wetland’s ecological integrity, because these species provide the habitats for many small plants and animals. In this study, we used hyperspectral imagery to map the distributions of two dominant tall grasses ( Miscanthus sacchariflorus (Maxim.) Benth and Phragmites australis (Cav.) Trin. ex Stend) in the Watarase wetland, in central Japan. Stepwise multiple linear regression analysis was applied to the hyperspectral data to predict the shoot density and biomass of the two grasses. The independent data sets included original reflectance, band ratios, significant components identified by principal components analysis (PCA), and significant components identified by decision boundary feature extraction (DBFE). The coefficient of determination ( R 2 ) and the root-mean-square error (RMSE) of model calibration and validation were used to evaluate the models. The significant DBFE components showed better ability at predicting shoot density of the two grasses than the other variables in the validating areas. The RMSE values were 7.40/m 2 for M. sacchariflorus and 13.09/m 2 for P. australis, which amounted to errors of around 10.0% and 12.6%, respectively, of the maximum shoot density measured during our surveys. All variables showed similar performance at predicting biomass, but the results were less accurate than those for shoot density. Considering the performance of the DBFE components for both shoot density and biomass prediction, we suggest that these are the best indicators for estimating the abundance of the two grasses.

Kernel Nonparametric Weighted Feature Extraction for Hyperspectral Image Classification

In recent years, many studies show that kernel methods are computationally efficient, robust, and stable for pattern analysis. Many kernel-based classifiers were designed and applied to classify remote-sensed data, and some results show that kernel-based classifiers have satisfying performances. Many studies about hyperspectral image classification also show that nonparametric weighted feature extraction (NWFE) is a powerful tool for extracting hyperspectral image features. However, NWFE is still based on linear transformation. In this paper, the kernel method is applied to extend NWFE to kernel-based NWFE (KNWFE). The new KNWFE possesses the advantages of both linear and nonlinear transformation, and the experimental results show that KNWFE outperforms NWFE, decision-boundary feature extraction, independent component analysis, kernel-based principal component analysis, and generalized discriminant analysis.

Feature‐selection ability of the decision‐tree algorithm and the impact of feature‐selection/extraction on decision‐tree results based on hyperspectral data

The decision‐tree (DT) algorithm is a very popular and efficient data‐mining technique. It is non‐parametric and computationally fast. Besides forming interpretable classification rules, it can select features on its own. In this article, the feature‐selection ability of DT and the impacts of feature‐selection/extraction on DT with different training sample sizes were studied by using AVIRIS hyperspcetral data. DT was compared with three other feature‐selection methods; the results indicated that DT was an unstable feature selector, and the number of features selected by DT was strongly related to the sample size. Trees derived with and without feature‐selection/extraction were compared. It was demonstrated that the impacts of feature selection on DT were shown mainly as a significant increase in the number of tree nodes (14.13–23.81%) and moderate increase in tree accuracy (3.5–4.8%). Feature extraction, like Non‐parametric Weighted Feature Extraction (NWFE) and Decision Boundary Feature Extraction (DBFE), could enhance tree accuracy more obviously (4.78–6.15%) and meanwhile a decrease in the number of tree nodes (6.89–16.81%). When the training sample size was small, feature‐selection/extraction could increase the accuracy more dramatically (6.90–15.66%) without increasing tree nodes.

Dimensionality Reduction Based on Clonal Selection for Hyperspectral Imagery

A new stochastic search strategy inspired by the clonal selection theory in an artificial immune system is proposed for dimensionality reduction of hyperspectral remote-sensing imagery. The clonal selection theory is employed to describe the basic features of an immune response to an antigenic stimulus in order to meet the requirement of diversity in the antibody population. In our proposed strategy, dimensionality reduction is formulated as an optimization problem that searches an optimum with less number of features in a feature space. In line with this novel strategy, a feature subset search algorithm, clonal selection Feature-Selection (CSFS) algorithm, and a feature-weighting algorithm, Clonal-Selection Feature-Weighting (CSFW) algorithm, have been developed. In the CSFS, each solution is evolved in binary space, and the value of each bit is either 0 or 1, which indicates that the corresponding feature is either removed or selected, respectively. In CSFW, each antibody is directly represented by a string consisting of integer numbers and their corresponding weights. These algorithms are compared with the following four well-known algorithms: sequential forward selection, sequential forward floating selection, genetic-algorithm-based feature selection, and decision-boundary feature extraction using the hyperspectral remote-sensing imagery acquired by the Pushbroom Hyperspectral Imager and the Airborne Visible/Infrared Imaging Spectrometer, respectively. Experimental results demonstrate that CSFS and CSFW outperform other algorithms and hence provide effective new options for dimensionality reduction of hyperspectral remote-sensing imagery.

Fast Hyperspectral Feature Reduction Using Piecewise Constant Function Approximations

The high number of spectral bands that are obtained from hyperspectral sensors, combined with the often limited ground truth, solicits some kind of feature reduction when attempting supervised classification. This letter demonstrates that an optimal constant function representation of hyperspectral signature curves in the mean square sense is capable of representing the data sufficiently to outperform, or match, other feature reduction methods such as principal components transform, sequential forward selection, and decision boundary feature extraction for classification purposes on all of the four hyperspectral data sets that we have tested. The simple averaging of spectral bands makes the resulting features directly interpretable in a physical sense. Using an efficient dynamic programming algorithm, the proposed method can be considered fast.

Classification and Extraction of Spatial Features in Urban Areas Using High-Resolution Multispectral Imagery

Classification and extraction of spatial features are investigated in urban areas from high spatial resolution multispectral imagery. The proposed approach consists of three steps. First, as an extension of our previous work [pixel shape index (PSI)], a structural feature set (SFS) is proposed to extract the statistical features of the direction-lines histogram. Second, some methods of dimension reduction, including independent component analysis, decision boundary feature extraction, and the similarity-index feature selection, are implemented for the proposed SFS to reduce information redundancy. Third, four classifiers, the maximum-likelihood classifier, backpropagation neural network, probability neural network based on expectation-maximization training, and support vector machine, are compared to assess SFS and other spatial feature sets. We evaluate the proposed approach on two QuickBird datasets, and the results show that the new set of reduced spatial features has better performance than the existing length-width extraction algorithm and PSI

Comparison between several feature extraction/classification methods for mapping complicated agricultural land use patches using airborne hyperspectral data

Airborne hyperspectral remote sensing was applied to agricultural land in the Miura Peninsula, near the metropolis of Tokyo in Japan. The study area is characterized by complicated land use patches, which is the general characteristic of most agricultural lands in Japan. Several feature extraction/classification methods were examined in classifying the land use and plant species. The results showed that decision boundary feature extraction (DBFE) was better than principal component analysis (PCA) as the feature extraction method. Moreover, the pre‐classification process using NDVI that separates the whole study area into vegetated area and non‐vegetated areas also improved the classification accuracy. After the pre‐procedures, the land use and plant species were finally mapped by maximum likelihood classification (MLC) or extraction and classification of homogeneous objects (ECHO). The best kappa (overall accuracy) of classification was 0.914 (92.4%) and 0.924 (93.3%) for MLC and ECHO, respectively. The best accuracies of each category for the image were 79.5% to 100% for plant species (watermelon, pumpkin, marigold, grass and tree), 88.7% to 100% for soil types, 97.8% for concrete, and 99.4% for vinyl‐mulches. Although, built‐up area has low estimation accuracy, this did not affect the overall classification accuracy because it covers only a very small area.

Classification and feature extraction for remote sensing images from urban areas based on morphological transformations

Classification of panchromatic high-resolution data from urban areas using morphological and neural approaches is investigated. The proposed approach is based on three steps. First, the composition of geodesic opening and closing operations of different sizes is used in order to build a differential morphological profile that records image structural information. Although, the original panchromatic image only has one data channel, the use of the composition operations will give many additional channels, which may contain redundancies. Therefore, feature extraction or feature selection is applied in the second step. Both discriminant analysis feature extraction and decision boundary feature extraction are investigated in the second step along with a simple feature selection based on picking the largest indexes of the differential morphological profiles. Third, a neural network is used to classify the features from the second step. The proposed approach is applied in experiments on high-resolution Indian Remote Sensing 1C (IRS-1C) and IKONOS remote sensing data from urban areas. In experiments, the proposed method performs well in terms of classification accuracies. It is seen that relatively few features are needed to achieve the same classification accuracies as in the original feature space.

Lossy compression of hyperspectral data using vector quantization

Efficient compression techniques are required for the coding of hyperspectral data. Lossless compression is required in the transmission and storage of data within the distribution .system. Lossy techniques have a role in the initial analysis of hyperspectral data where large quantities of data are evaluated to select smaller areas for more detailed evaluation. Central to lossy compression is the developmgent of a suitable distortion measure, and this work discusses the applicability of extant measures in video coding to the compression of hyperspectral imagery. Criteria for a remote sensing distortion measure are developed and suitable distortion measures are discussed. One measure [the percentage maximum absolute distortion (PMAD) measure] is considered to be a suitable candidate for application to remotely sensed images. The effect of lossy compression is then investigated on the maximum likelihood classification of hyperspectral images, both directly on the original reconstructed data and on features extracted by the decision boundary feature extraction (DBFE) technique. The effect of the PMAD measure is determined on the classification of an image reconstructed with varying degrees of distortion. Despite some anomalies caused by challenging discrimination tasks, the classification accuracy of both the total image and its constituent classes remains predictable as the level of distortion increases. Although total classification accuracy is reduced from 96.8% for the original image to 82.8% for the image compressed with 4% PMAD, the loss in accuracy is not significant (less that 8%)for most classes other than those that present a challenging classification problem. Yet the compressed image is 1/17 the size of the original.

Feature extraction for multisource data classification with artificial neural networks

Classification of multisource remote sensing and geographic data by neural networks is discussed with respect to feature extraction. Several feature extraction methods are reviewed, including principal component analysis, discriminant analysis, and the recently proposed decision boundary feature extraction method. The feature extraction methods are then applied in experiments in conjunction with classification by multilayer neural networks. The decision boundary feature extraction method shows excellent performance in the experiments.