By Topic

Selected Topics in Applied Earth Observations and Remote Sensing, IEEE Journal of

Issue 6 • Date June 2014

Filter Results

Displaying Results 1 - 25 of 80
  • Front cover

    Page(s): C1
    Save to Project icon | Request Permissions | PDF file iconPDF (281 KB)  
    Freely Available from IEEE
  • IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing publication information

    Page(s): C2
    Save to Project icon | Request Permissions | PDF file iconPDF (137 KB)  
    Freely Available from IEEE
  • Table of contents

    Page(s): 1837 - 1840
    Save to Project icon | Request Permissions | PDF file iconPDF (148 KB)  
    Freely Available from IEEE
  • Foreword to the Special Issue on Hyperspectral Image and Signal Processing

    Page(s): 1841 - 1843
    Save to Project icon | Request Permissions | PDF file iconPDF (155 KB)  
    Freely Available from IEEE
  • A Review of Nonlinear Hyperspectral Unmixing Methods

    Article#: 2320576
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1665 KB) |  | HTML iconHTML  

    In hyperspectral unmixing, the prevalent model used is the linear mixing model, and a large variety of techniques based on this model has been proposed to obtain endmembers and their abundances in hyperspectral imagery. However, it has been known for some time that nonlinear spectral mixing effects can be a crucial component in many real-world scenarios, such as planetary remote sensing, intimate mineral mixtures, vegetation canopies, or urban scenes. While several nonlinear mixing models have been proposed decades ago, only recently there has been a proliferation of nonlinear unmixing models and techniques in the signal processing literature. This paper aims to give an historical overview of the majority of nonlinear mixing models and nonlinear unmixing methods, and to explain some of the more popular techniques in detail. The main models and techniques treated are bilinear models, models for intimate mineral mixtures, radiosity-based approaches, ray tracing, neural networks, kernel methods, support vector machine techniques, manifold learning methods, piece-wise linear techniques, and detection methods for nonlinearity. Furthermore, we provide an overview of several recent developments in the nonlinear unmixing literature that do not belong into any of these categories. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A Comparison of Nonlinear Mixing Models for Vegetated Areas Using Simulated and Real Hyperspectral Data

    Article#: 2328872
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1670 KB) |  | HTML iconHTML  

    Spectral unmixing (SU) is a crucial processing step when analyzing hyperspectral data. In such analysis, most of the work in the literature relies on the widely acknowledged linear mixing model to describe the observed pixels. Unfortunately, this model has been shown to be of limited interest for specific scenes, in particular when acquired over vegetated areas. Consequently, in the past few years, several nonlinear mixing models have been introduced to take nonlinear effects into account while performing SU. These models have been proposed empirically, however, without any thorough validation. In this paper, the authors take advantage of two sets of real and physical-based simulated data to validate the accuracy of various nonlinear models in vegetated areas. These physics-based models, and their corresponding unmixing algorithms, are evaluated with respect to their ability of fitting the measured spectra and providing an accurate estimation of the abundance coefficients, considered as the spatial distribution of the materials in each pixel. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A Distance Geometric Framework for Nonlinear Hyperspectral Unmixing

    Article#: 2319894
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1306 KB) |  | HTML iconHTML  

    In this article, a distance geometry-based framework for hyperspectral image unmixing is presented. A manifold representation of the data set is generated by creation of a nearest-neighbor graph on which shortest paths are calculated yielding a geodesic distance matrix. Instead of unfolding the manifold in a lower-dimensional Euclidean space, it is proposed to work directly on the manifold. To do so, algorithms need to be rewritten in terms of distance geometry. Building further on earlier work, where distance-based dimensionality estimation and endmember extraction methods were presented, we will propose a distance geometric version of the actual unmixing (abundance estimation) step. In this way, a complete distance geometric unmixing framework is obtained that is efficient compared to the classical methods based on optimization. Furthermore, the distance geometry-adapted algorithms can be applied on nonlinear data manifolds by employing geodesic distances. In the experiments, we demonstrate this by comparing the obtained nonlinear framework to its linear counterpart. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Non-Local Sparse Unmixing for Hyperspectral Remote Sensing Imagery

    Page(s): 1889 - 1909
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (6041 KB) |  | HTML iconHTML  

    Sparse unmixing is a promising approach that acts as a semi-supervised unmixing strategy by assuming that the observed image signatures can be expressed in the form of linear combinations of a number of pure spectral signatures that are known in advance. However, conventional sparse unmixing involves finding the optimal subset of signatures for the observed data in a very large standard spectral library, without considering the spatial information. In this paper, a new sparse unmixing algorithm based on non-local means, namely non-local sparse unmixing (NLSU), is proposed to perform the unmixing task for hyperspectral remote sensing imagery. In NLSU, the non-local means method, as a regularizer for sparse unmixing, is used to exploit the similar patterns and structures in the abundance image. The NLSU algorithm based on the sparse spectral unmixing model can improve the spectral unmixing accuracy by incorporating the non-local spatial information by means of a weighting average for all the pixels in the abundance image. Five experiments with three simulated and two real hyperspectral images were performed to evaluate the performance of the proposed algorithm in comparison to the previous sparse unmixing methods: sparse unmixing via variable splitting and augmented Lagrangian (SUnSAL) and sparse unmixing via variable splitting augmented Lagrangian and total variation (SUnSAL-TV). The experimental results demonstrate that NLSU outperforms the other algorithms, with a better spectral unmixing accuracy, and is an effective spectral unmixing algorithm for hyperspectral remote sensing imagery. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Earth Movers Distance-Based Simultaneous Comparison of Hyperspectral Endmembers and Proportions

    Page(s): 1910 - 1921
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (2071 KB) |  | HTML iconHTML  

    A new approach for simultaneously comparing sets of hyperspectral endmembers and proportion values using the Earth Movers Distance (EMD) is presented. First, the EMD is defined and calculated per-pixel based on the proportion values and corresponding endmembers. Next, these per-pixel EMD distances are aggregated to obtain a final measure of dissimilarity. In particular, the proposed EMD approach can be used to simultaneously compare endmembers and proportion values with differing numbers of endmembers. The proposed method has a number of uses, including: computing the similarity between two sets of endmembers and proportion values that were obtained using any algorithm or underlying mixing model, clustering sets of hyperspectral endmember and proportion values, or evaluating spectral unmixing results by comparing estimated values to ground truth information. Experiments on both simulated and measured hyperspectral data sets demonstrate that the EMD is effective at simultaneous endmember and proportion comparison. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Unsupervised Spectral Mixture Analysis of Highly Mixed Data With Hopfield Neural Network

    Page(s): 1922 - 1935
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (2681 KB) |  | HTML iconHTML  

    Hopfield Neural Network (HNN) has been demonstrated to be an effective tool for Spectral Mixture Analysis (SMA). However, the spectrum of pure ground objects, known as endmember, must be known previously. In this paper, the HNN is utilized to solve unsupervised SMA, in which Endmember Extraction (EE) and Abundance Estimation (AE) are performed iteratively. Two different HNNs are constructed to solve such multiplicative updating procedure, respectively. The proposed HNN based unsupervised SMA framework is then applied to solve three second-order constrained Nonnegative Matrix Factorization (NMF) models for SMA, including Minimum Distance Constrained NMF (MDC-NMF), Minimum endmember-wise Distance Constrained NMF (MewDC-NMF), and Minimum Dispersion Constrained NMF (MiniDisCo-NMF). As a result, our proposed HNN based algorithms are able to perform unsupervised SMA and extract virtual endmembers without assuming the presence of spectrally pure constituents in highly mixed hyperspectral data. Experimental results on both synthetic and real hyperspectral images demonstrate that our proposed HNN based algorithms clearly outperform traditional Projected Gradient (PG) based solutions for these constrained NMF based SMA. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A Data-Driven Stochastic Approach for Unmixing Hyperspectral Imagery

    Article#: 2328597
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (2210 KB) |  | HTML iconHTML  

    In this paper, we propose a two-step Bayesian approach to handle the ill-posed nature of the unmixing problem for accurately estimating the abundances. The abundances are dependent on the scene contents and they represent mixing proportions of the endmembers over an area. In this work, a linear mixing model (LMM) is used for the image formation process in order to derive the data term. In the first step, a Huber-Markov random field (HMRF)-based prior distribution is assumed to model the dependencies within the abundances across the spectral space of the data. The threshold used in the HMRF prior is derived from an initial estimate of abundances obtained using the matched filters. This makes the HMRF prior data-driven, i.e., dHMRF. Final abundance maps are obtained in the second step within a maximum a posteriori probability (MAP) framework, and the objective function is optimized using the particle swarm optimization (PSO). Theoretical analysis is carried out to show the effectiveness of the proposed method. The approach is evaluated using the synthetic and real AVIRIS Cuprite data. The proposed method has the following advantages. 1) The estimated abundances are resistant to noise since they are based on an initial estimate that has high signal-to-noise ratio (SNR). 2) The variance in the abundance maps is well preserved since the threshold in the dHMRF is derived from the data. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Fuzzy Assessment of Spectral Unmixing Algorithms

    Page(s): 1947 - 1955
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1501 KB) |  | HTML iconHTML  

    While a single spectrum is often used to present a pure class, it is more realistic to consider intra-class spectral variation and model a pure class using a group of its representative spectra. In line with this consideration, crisp unmixing accuracy assessment, where unmixing performance is assessed using a mean squared error of the estimated endmember fractions, can be misleading. In this paper, alterative spectral unmixing assessment methods are introduced to account for the uncertainty contained in the spectral measurements and during the ground truth data collection. Two fuzzy measures are developed to assess unmixing performance. One is fuzzy unmixing fraction error for a realistic assessment and the other is pixel level unmixing accuracy to provide a good quantitative understanding of the unmixing success rates spatially. To demonstrate and illustrate how they work, the two fuzzy measures are applied to evaluate the performance of several spectral unmixing methods including both single spectrum based and multiple spectra based algorithms. Crisp assessments and fuzzy results at various tolerance levels are presented and compared. Based on the realistic measures proposed, it is found the recent developed unmixing method with extended Support Vector Machines outperforms other algorithms tested. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Quantifying Nonlinear Spectral Mixing in Vegetated Areas: Computer Simulation Model Validation and First Results

    Page(s): 1956 - 1965
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1506 KB) |  | HTML iconHTML  

    Our understanding of nonlinear mixing events in vegetated areas is currently hampered by a pertinent lack of well-validated datasets. Most quantification and modeling efforts are, therefore, based on the theoretical assumptions or indirect empirical observations. Here, we performed a quantitative and qualitative evaluation of the accuracy of nonlinear mixing effects as modeled by a fully calibrated virtual orchard model (based on physically based ray-tracer software). For validation, we had available data from an in situ experiment. This experiment comprised in situ measured mixed pixel reflectance spectra, pixel-specific endmember spectra, and subpixel cover fraction distributions, all collected in the same orchard for which the virtual model was calibrated. We took advantage of this unique-coupled dataset to demonstrate that both the nature and the intensity of the nonlinear mixing events observed in the in situ data are realistically modeled by the ray-tracing software. This is an important observation because this implies that our virtual model now provides a solid tool for the detailed study of nonlinear mixing in vegetated areas which could facilitate as such the design, calibration, and validation of different nonlinear mixing modeling approaches. Initial results revealed that the nonlinear mixing is dependent on fractional distribution, soil moisture conditions, and endmember definitions. We could further demonstrate that the bilinear spectral mixture model nicely described nonlinear mixing events but at the same time overestimated reflectances in spectral regions with moderate-to-low nonlinear mixing behavior. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Subspace-Projection-Based Geometric Unmixing for Material Quantification in Hyperspectral Imagery

    Article#: 2328581
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1617 KB) |  | HTML iconHTML  

    Linear spectral unmixing is a widely used technique in hyperspectral remote sensing to quantify materials present in an image pixel. In order to produce accurate estimates of abundances, nonnegativity constraint and sum-to-one constraint must be imposed on the abundances of materials. Under these two constraints, linear spectral unmixing is often formulated as a convex optimization problem that requires more advanced optimization technology, leading to excessive computational complexity. In this paper, a novel geometric method is presented for solving the fully constrained linear spectral unmixing problem. Specifically, abundances are first expressed as the ratios of signed volumes of simplexes. Then, Laplace expansion is applied in the process of determinant calculation, which derives a new low-complexity abundance estimation method. Furthermore, the mixed pixel outside the simplex is iteratively projected onto the facet planes through the endmember vertices for making the abundances satisfy the nonnegativity constraint. This process is continued until one finds a projected point lying inside the simplex. The proposed method is in line with the least squares criterion. Experimental results based on simulated and the AVIRIS Cuprite data sets demonstrate the superiority of the proposed algorithm with respect to other state-of-the-art approaches. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Spatially Constrained Multiple Endmember Spectral Mixture Analysis for Quantifying Subpixel Urban Impervious Surfaces

    Article#: 2318018
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1450 KB) |  | HTML iconHTML  

    Multiple endmember spectral mixture analysis (MESMA) has been extensively employed to accommodate endmember variability associated with the mixed pixel problem in remote sensing imagery. However, endmember extraction is a critical step in the application of MESMA. Considering that spatial information can be helpful for selecting local representative endmembers, this paper develops a spatially constrained MESMA method, with which multiple endmembers for each class are automatically derived within a predefined neighborhood. Two specific novelties are: 1) to identify all the endmembers over the whole image scene for each class through a classification tree approach; and 2) to generate spatially constrained endmembers for the neighborhood of each target pixel of the image through a k-means clustering method. MESMA is then performed using the derived spatially constrained endmembers. This proposed method was applied to a Landsat Enhanced Thematic Mapper (ETM+) image for examining subpixel urban impervious surfaces, and its performance was compared with that of a global MESMA method. The results suggest that spatially constrained MESMA is able to yield adequate estimates, supported by a relatively decent precision and low bias (10.68% for mean absolute error and -3.58% for systematic error). View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Integrating Spatial Information in Unsupervised Unmixing of Hyperspectral Imagery Using Multiscale Representation

    Article#: 2319261
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1827 KB) |  | HTML iconHTML  

    This paper presents an unsupervised unmixing approach that takes advantage of multiscale representation based on nonlinear diffusion to integrate the spatial information in the spectral endmembers extraction from a hyperspectral image. The main advantages of unsupervised unmixing based on multiscale representation (UUMR) are the avoidance of matrix rank estimation to determine the number of endmembers and the use of spatial information without employing spatial kernels. Multiscale representation builds a family of smoothed images where locally spectrally uniform regions can be identified. The multiscale representation is extracted solving a nonlinear diffusion partial differential equation (PDE). Locally, homogeneous regions are identified by taking advantage of an algebraic multigrid method used to solve the PDE. Representative spectra for each region are extracted and then clustered to build spectral endmember classes. These classes represent the different spectral components of the image as well as their spectral variability. The number of spectral endmember classes is estimated using the Davies and Bouldin validity index. A quantitative assessment of unmixing approach based on multiscale representation is presented using an AVIRIS image captured over Fort. A.P. Hill, Virginia. A comparison of UUMR results with others unmixing techniques is included. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Spatial and Spectral Unmixing Using the Beta Compositional Model

    Article#: 2330347
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1995 KB) |  | HTML iconHTML  

    This paper introduces the beta compositional model (BCM) for hyperspectral unmixing and four algorithms for unmixing given the BCM. Hyperspectral unmixing estimates the proportion of each endmember at every pixel of a hyperspectral image. Under the BCM, each endmember is a random variable distributed according to a beta distribution. By using a beta distribution, spectral variability is accounted for during unmixing, the reflectance values of each endmember are constrained to a physically realistic range, and skew can be accounted for in the distribution. Spectral variability is incorporated to increase hyperspectral unmixing accuracy. Two BCM-based spectral unmixing approaches are presented: BCM-spectral and BCM-spatial. For each approach, two algorithms, one based on quadratic programming (QP) and one using a Metropolis-Hastings (MH) sampler, are developed. Results indicate that the proposed BCM unmixing algorithms are able to successfully perform unmixing on simulated data and real hyperspectral imagery while incorporating endmember spectral variability and spatial information. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Spatial-Spectral Information Based Abundance-Constrained Endmember Extraction Methods

    Page(s): 2004 - 2015
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1597 KB) |  | HTML iconHTML  

    Endmember extraction, which is an important technique for hyperspectral data interpretation, selects a collection of pure signature spectra of the different materials, called endmembers, which are present in a remotely sensed hyperspectral image scene. These pure signatures are then used in spectral unmixing algorithms to decompose the scene into abundance fractions, which indicate the proportion of each endmember's presence in a mixed pixel. In other words, abundances can be obtained by the given endmembers. Correspondingly, endmembers can be extracted based on an abundance constraint. In this paper, we first propose an endmember extraction framework based on an abundance constraint whose efficiency is related to the abundance calculation. The mainstream existing spatial-spectral algorithms can have a very high complexity and are sensitive to outliers, or the spatial information is considered followed by the spectral information. We therefore propose a strategy to consider the spectral information followed by the spatial information, using an abundance-constrained framework. The spatial strategy is also assumed to be immune to outliers. Experiments on both synthetic and real hyperspectral data sets indicate that: 1) the abundance constraint is effective for endmember extraction; and 2) the proposed spatial processing method used in the abundance-constrained endmember extraction framework can effectively avoid outliers. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A Dynamic Unmixing Framework for Plant Production System Monitoring

    Article#: 2314960
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (3890 KB) |  | HTML iconHTML  

    Hyperspectral remote sensing or imaging spectroscopy is an emerging technology in plant production monitoring and management. The continuous reflectance spectra allow for the intensive monitoring of biophysical and biochemical tree characteristics during growth, through for instance the use of vegetation indices. Yet, since most of the pixels in hyperspectral images are mixed, the evaluation of the actual vegetation state on the ground directly from the measured spectra is degraded by the presence of other endmembers, such as soil. Spectral unmixing, then, becomes a necessary processing step to improve the interpretation of vegetation indices. In this sense, an active research direction is based on the use of large collections of pure spectra, called spectral libraries or dictionaries, which model a wide variety of possible states of the endmembers of interest on the ground, i.e., vegetation and soil. Under the linear mixing model (LMM), the observed spectra are assumed to be linear combinations of spectra from the available dictionary. Combinatorial techniques (e.g., MESMA) and sparse regression algorithms (e.g., SUnSAL) are widely used to tackle the unmixing problem in this case. However, both combinatorial and sparse techniques benefit from appropriate library reduction strategies. In this paper, we develop a new efficient method for library reduction (or dictionary pruning), which exploits the fact that hyperspectral data generally lives in a lower-dimensional subspace. Specifically, we present a slight modification of the MUSIC-CSR algorithm, a two-step method which aims first at pruning the dictionary and second at infering high-quality reconstruction of the vegetation spectra on the ground (this application being called signal unmixing in remote sensing), using the pruned dictionary as input to available unmixing methods. Our goal is two-fold: 1) to obtain high-accuracy unmixing output using sparse unmixing, with low-execution time; and 2) to improve MESMA p- rformances in terms of accuracy. Our experiments, which have been conducted in a multi-temporal case study, show that the method achieves these two goals and proposes sparse unmixing as a reliable and robust alternative to the combinatorial methods in plant production monitoring applications. We further demonstrate that the proposed methodology of combining a library pruning approach with spectral unmixing provides a solid framework for the year-round monitoring of plant production systems. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Spectral-Spatial Classification of Hyperspectral Image Based on Discriminant Analysis

    Page(s): 2035 - 2043
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (2480 KB) |  | HTML iconHTML  

    This paper proposes a spectral-spatial linear discriminant analysis (LDA) method for the hyperspectral image classification. A natural assumption is that similar samples have similar structure in the dimensionality reduced feature space. The proposed method uses a local scatter matrix from a small neighborhood as a regularizer incorporated into the objective function of LDA. Different from traditional LDA and its variants, our proposed method yields a self-adaptive projection matrix for dimension reduction, which improves the classification accuracy and avoids running out of memory. In order to consider the nonlinear case, this paper generalizes our linear version to its kernel version. Experimental results demonstrate that our proposed methods outperform several dimension reduction algorithms. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Modified Co-Training With Spectral and Spatial Views for Semisupervised Hyperspectral Image Classification

    Article#: 2325741
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (2188 KB) |  | HTML iconHTML  

    Hyperspectral images are characterized by limited labeled samples, large number of spectral channels, and existence of noise and redundancy. Supervised hyperspectral image classification is difficult due to the unbalance between the high dimensionality of the data and the limited labeled training samples available in real analysis scenarios. The collection of labeled samples is generally hard, expensive, and time-consuming, whereas unlabeled samples can be obtained much easier. This observation has fostered the idea of adopting semisupervised learning techniques in hyperspectral image classification. In this paper, a semisupervised method based on a modified co-training process with spectral and spatial views is proposed for hyperspectral image classification. The original spectral features and the 2-D Gabor features extracted from spatial domains are adopted as two distinct views for co-training, which considers both the spectral and spatial information. Then, a modified co-training process with a new sample selection scheme is presented, which can effectively improve the co-training performance, especially when there are extremely limited labeled samples available. Experiments carried out on two real hyperspectral images show the superiority of the proposed semisupervised method with the modified co-training process over the corresponding supervised techniques, the semisupervised method with the conventional co-training version, and the semisupervised graph-based method. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A Nonlocal Weighted Joint Sparse Representation Classification Method for Hyperspectral Imagery

    Page(s): 2056 - 2065
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (2125 KB) |  | HTML iconHTML  

    As a powerful and promising statistical signal modeling technique, sparse representation has been widely used in various image processing and analysis fields. For hyperspectral image classification, previous studies have shown the effectiveness of the sparsity-based classification methods. In this paper, we propose a nonlocal weighted joint sparse representation classification (NLW-JSRC) method to improve the hyperspectral image classification result. In the joint sparsity model (JSM), different weights are utilized for different neighboring pixels around the central test pixel. The weight of one specific neighboring pixel is determined by the structural similarity between the neighboring pixel and the central test pixel, which is referred to as a nonlocal weighting scheme. In this paper, the simultaneous orthogonal matching pursuit technique is used to solve the nonlocal weighted joint sparsity model (NLW-JSM). The proposed classification algorithm was tested on three hyperspectral images. The experimental results suggest that the proposed algorithm performs better than the other sparsity-based algorithms and the classical support vector machine hyperspectral classifier. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Classification of Hyperspectral Data Using an AdaBoostSVM Technique Applied on Band Clusters

    Page(s): 2066 - 2079
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (2409 KB)  

    Supervised classification of hyperspectral image data using conventional statistical classification methods is difficult because a sufficient number of training samples is often not available for the wide range of spectral bands. In addition, spectral bands are usually highly correlated and contain data redundancies because of the short spectral distance between the adjacent bands. To address these limitations, a multiple classifier system based on Adaptive Boosting (AdaBoost) is proposed and evaluated to classify hyperspectral data. In this method, the hyperspectral datasets are first split into several band clusters based on the similarities between the contiguous bands. In an AdaBoost classification system, the redundant and noninformative bands in each cluster are then removed using an optimal band selection technique. Next, a support vector machine (SVM) is applied to each refined cluster based on the classification results of previous clusters, and the results of these classifiers are fused using the weights obtained from the AdaBoost processing. Experimental results with standard hyperspectral datasets clearly demonstrate the superiority of the proposed algorithm with respect to both global and class accuracies, when compared to another ensemble classifiers such as simple majority voting and Naïve Bayes to combine decisions from each cluster, a standard SVM applied on the selected bands of entire datasets and on all the spectral bands. More specifically, the proposed method performs better than other approaches, especially in datasets which contain classes with greater complexity and fewer available training samples. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Dynamic Linear Classifier System for Hyperspectral Image Classification for Land Cover Mapping

    Page(s): 2080 - 2093
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (3912 KB) |  | HTML iconHTML  

    Exploitation of the spectral capabilities of modern hyperspectral image demands efficient preprocessing and analyses methods. Analysts' choice of classifier and dimensionality reduction (DR) method and the harmony between them determine the accuracy of image classification. Multiple classifier system (MCS) has the potential to combine the relative advantages of several classifiers into a single image classification exercise for the hyperspectral image classification. In this paper, we propose an algorithmic extension of the MCS, named as dynamic classifier system (DCS), which exploits the context-based image and information class characteristics represented by multiple DR methods for hyperspectral image classification for land cover mapping. The proposed DCS algorithm pairs up optimal combinations of classifiers and DR methods specific to the hyperspectral image and performs image classifications based only on the identified combinations. Further, the impact of various trainable and nontrainable combination functions on the performance of the proposed DCS has been assessed. Image classifications were carried out on five multi-site airborne hyperspectral images using the proposed DCS and were compared with the MCS and SVM based supervised image classifications with and without DR. The results indicate the potential of the proposed DCS algorithm to increase the classification accuracy considerably over that of MCS or SVM supervised image classifications. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Deep Learning-Based Classification of Hyperspectral Data

    Article#: 2329330
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (2730 KB) |  | HTML iconHTML  

    Classification is one of the most popular topics in hyperspectral remote sensing. In the last two decades, a huge number of methods were proposed to deal with the hyperspectral data classification problem. However, most of them do not hierarchically extract deep features. In this paper, the concept of deep learning is introduced into hyperspectral data classification for the first time. First, we verify the eligibility of stacked autoencoders by following classical spectral information-based classification. Second, a new way of classifying with spatial-dominated information is proposed. We then propose a novel deep learning framework to merge the two features, from which we can get the highest classification accuracy. The framework is a hybrid of principle component analysis (PCA), deep learning architecture, and logistic regression. Specifically, as a deep learning architecture, stacked autoencoders are aimed to get useful high-level features. Experimental results with widely-used hyperspectral data indicate that classifiers built in this deep learning-based framework provide competitive performance. In addition, the proposed joint spectral-spatial deep neural network opens a new window for future research, showcasing the deep learning-based methods' huge potential for accurate hyperspectral data classification. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.

Aims & Scope

IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing (J-STARS) addresses current issues and techniques in applied remote and in situ sensing, their integration, and applied modeling and information creation for understanding the Earth.

Full Aims & Scope

Meet Our Editors

Editor-in-Chief
Dr. Jocelyn Chanussot
Grenoble Institute of Technology