Hyperspectral Digital Imagery Collection Experiment Research Articles

FPGA Implementation of a Hardware Optimized Automatic Target Detection and Classification Algorithm for Hyperspectral Image Analysis

In hyperspectral image analysis one of the most important tasks is target detection, requiring the execution of algorithms with high computational complexity. Recently, research efforts have focused on on-board real-time target detection to provide timely responses for swift decisions. Therefore, it is necessary to use a technology that provides the performance needed for real-time target detection, and at the same time meets the satellite payload requirements. Field-programmable gate arrays (FPGAs) have very interesting properties in terms of performance, size and power consumption, which have become the standard option for on-board processing. In this letter, we present a hardware optimized implementation for FPGAs of the automatic target detection and classification algorithm (ATDCA) using the Gram–Schmidt (GS) method for orthogonalization purposes. The ATDCA-GS algorithm is directly coded using VHDL and verified on a Virtex-7 XC7VX690T FPGA using real hyperspectral data (collected by Hyperspectral Digital Imagery Collection Experiment (HYDICE) sensor and by NASA’s Airborne Visible/Infrared Imaging Spectrometer (AVIRIS)) and a synthetic image. Experimental results demonstrate that our hardware version of the ATDCA-GS algorithm outperforms previous implementations (multicore processors, GPUs and accelerators) in both computation time (obtaining real-time performance) and power consumption, demonstrating the suitability of FPGAs for this purpose.

Geometrical constrained independent component analysis for hyperspectral unmixing

ABSTRACT One of the limitations of the hyperspectral remote sensing application is the existence of mixed pixels in image data. Spectral decomposition, which separates the mixed pixels into a set of endmember spectra and abundance fractions, is the most effective way to solve the mixed pixel problem. Due to the independence of source signals, independent component analysis (ICA) has been developed for hyperspectral unmixing by adding auxiliary constraints. Abundance sum-to-one and nonnegative constraints are two obvious features for hyperspectral data. In this paper, by processing these two constraints sequentially from the geometrical point of view to restrain the sum-to-one constraint thoroughly at each iteration, geometrical constrained ICA (GCICA) is proposed based on treating the abundance distribution as the independent signal. To validate the proposed algorithm, the synthetic data, and real image data are used for unmixing, respectively. Synthetic data are generated based on spectra from the ENVI (Environment for Visualizing Images) software spectral library. The real images are used with three hyperspectral datasets, AVIRIS (Airborne Visible Infrared Imaging Spectrometer) Cuprite dataset, AVIRIS Indiana Pine dataset and HYDICE (Hyperspectral Digital Imagery Collection Experiment) dataset. Results, in comparison with previously proposed algorithms, show that the proposed method has better performance for the decomposition of hyperspectral data in abundance and endmember spectral extraction, thus providing a new and effective method for spectral unmixing and signal separation without prior information.

An improved cuckoo search-based adaptive band selection for hyperspectral image classification

ABSTRACT The information in hyperspectral images usually has a strong correlation, a large number of bands, which lead to the “curse of dimensionality”. So, band selection is usually used to address this issue. However, problems remain for band selection, such as how to search for the most informative bands, and how many bands should be selected. In this paper, a cuckoo search (CS)-based adaptive band selection framework is proposed to simultaneously select bands and determine the optimal number of bands to be selected. The proposed framework includes two “cuckoo search”, i.e. the outer one for estimating the optimal number of bands and the inner one for the corresponding band selection. To avoid employing an actual classifier within CS so as to greatly reduce computational cost, minimum estimated abundance covariance (MEAC) and Jeffreys-Matusita (JM) distance are adopted as criterion functions, which measures class separability. For the experiments, two widely used hyperspectral images, which acquired by the Hyperspectral digital imagery collection experiment (HYDICE) and the airborne Hyperspectral Mapper (HYMAP) system, are adopted for performance evaluation. The experimental results show that the two-CS-based algorithm outperforms the popular sequential forward selection (SFS), sequential floating forward search (SFFS), and other similar algorithms for hyperspectral band selection.

Portability Study of an OpenCL Algorithm for Automatic Target Detection in Hyperspectral Images

In the last decades, the problem of target detection has received considerable attention in remote sensing applications. When this problem is tackled using hyperspectral images with hundreds of bands, the use of high-performance computing (HPC) is essential. One of the most popular algorithms in the hyperspectral image analysis community for this purpose is the automatic target detection and classification algorithm (ATDCA). Previous research has already investigated the mapping of ATDCA on HPC platforms such as multicore processors, graphics processing units (GPUs), and field-programmable gate arrays (FPGAs), showing impressive speedup factors (after careful fine-tuning) that allow for its exploitation in time-critical scenarios. However, the lack of standardization resulted in most implementations being too specific to a given architecture, eliminating (or at least making extremely difficult) code reusability across different platforms. In order to address this issue, we present a portability study of an implementation of ATDCA developed using the open computing language (OpenCL). We focus on cross-platform parameters such as performance, energy consumption, and code design complexity, as compared to previously developed (hand-tuned) implementations. Our portability study analyzes different strategies to expose data parallelism as well as enable the efficient exploitation of complex memory hierarchies in heterogeneous devices. We also conduct an assessment of energy consumption and discuss metrics to analyze the quality of our code. The conducted experiments—using synthetic and real hyperspectral data sets collected by the Hyperspectral Digital Imagery Collection Experiment (HYDICE) and NASA’s Airborne Visible Infra-Red Imaging Spectrometer (AVIRIS)—demonstrate, for the first time in the literature, that portability across different HPC platforms can be achieved for real-time target detection in hyperspectral missions.

Classification of Urban Hyperspectral Remote Sensing Imagery Based on Optimized Spectral Angle Mapping

Hyperspectral remote sensing imagery provides highly precise spectral information. Thus, it is suitable for the land use classification of urban areas that are composed of complicated structures. In this study, a new spectral angle and vector mapping (SAVM) classification method, which adds a factor based on “the differences in the spectral vector lengths” among image pixels to the spectral angle mapping (SAM) classification method, is proposed. The SAM and SAVM methods were applied to classify the aerial hyperspectral digital imagery collection experiment imagery acquired from the business district of Washington, DC, USA. The results demonstrated that the overall classification accuracy of the SAM was 64.29%, with a Kappa coefficient of 0.57, while the overall classification accuracy of the SAVM was 81.06%, with a Kappa coefficient of 0.76. The overall classification accuracy was improved by 16.77% by the SAVM, indicating that the use of a SAVM classification method that considers both the spectral angle between the reference spectrum and the test spectrum and the differences in the spectral vector lengths among image pixels can improve the classification accuracy of urban area with hyperspectral remote sensing imagery.

CLASS-BASED AFFINITY PROPAGATION FOR HYPERSPECTRAL IMAGE DIMENSIONALITY REDUCTION AND IMPROVEMENT OF MAXIMUM LIKELIHOOD CLASSIFICATION ACCURACY

Abstract This paper investigates an alternative classification method that integrates class-based affinity propagation (CAP) clustering algorithm and maximum likelihood classifier (MLC) with the purpose of overcome the MLC limitations in the classification of high dimensionality data, and thus improve its accuracy. The new classifier was named CAP-MLC, and comprises two approaches, spectral feature selection and image classification. CAP clustering algorithm was used to perform the image dimensionality reduction and feature selection while the MLC was employed for image classification. The performance of MLC in terms of classification accuracy and processing time is determined as a function of the selection rate achieved in the CAP clustering stage. The performance of CAP-MLC has been evaluated and validated using two hyperspectral scenes from the Airborne Visible Infrared Imaging Spectrometer (AVIRIS) and the Hyperspectral Digital Imagery Collection Experiment (HYDICE). Classification results show that CAP-MLC observed an enormous improvement in accuracy, reaching 94.15% and 96.47% respectively for AVIRIS and HYDICE if compared with MLC, which had 85.42% and 81.50%. These values obtained by CAP-MLC improved the MLC classification accuracy in 8.73% and 14.97% for these images. The results also show that CAP-MLC performed well, even for classes with limited training samples, surpassing the limitations of MLC.

Dimensionality reduction based on binary encoding for hyperspectral data

ABSTRACTBinary encoding is an approach that aims at summarizing the information contained in various spectral bands into a single image that stores the meaningful information of the bands. In this paper, it is introduced a feature extraction approach to reduce the dimensionality of hyperspectral data with binary encoding for classification purposes. Different options to reduce the radiometric information of the pixels are introduced, such as using a single threshold or multiple thresholds. After the dimensionality reduction, the separation of the spectral classes was analysed and the thematic classification of the reduced data was performed. In order to evaluate the performance of the proposed approach, experiments on AVIRIS (Airborne Visible/Infrared Imaging Spectrometer) image, ROSIS (Reflection Optics System Imaging Spectrometer) hyperspectral image and HYDICE (Hyperspectral Digital Imagery Collection Experiment) hyperspectral image are presented. In the experiments, neighbouring spectral bands are grouped and coded and the results of the classification are compared. The results show that the use of binary encoding based on three thresholds by spectral region is more efficient than with the use of one threshold. The thematic mapping of the hyperspectral data with reduced dimension confirms the competitiveness of the binary encoding method compared with other dimension reduction methods, such as the Principal Component Analysis (PCA), the Principal Component Analysis – Fisher’s Linear Discriminant Analysis (PCA-LDA), the Discriminant Analysis Feature Extraction (DAFE) and the Non-parametric Weighted Feature Extraction (NWFE). In this context, the present methodology shows to be promising, because it reduces the computational complexity and improves performance.

Noise Removal Based on Tensor Modelling for Hyperspectral Image Classification

With the current state-of-the-art computer aided manufacturing tools, the spatial resolution of hyperspectral sensors is becoming increasingly higher thus making it easy to obtain much more detailed information of the scene captured. However, the improvement of the spatial resolution also brings new challenging problems to address with signal dependent photon noise being one of them. Unlike the signal independent thermal noise, the variance of photon noise is dependent on the signal, therefore many denoising methods developed for the stationary noise cannot be applied directly to the photon noise. To make things worse, both photon and thermal noise coexist in the captured hyperspectral image (HSI), thus making it more difficult to whiten noise. In this paper, we propose a new denoising framework to cope with signal dependent nonwhite noise (SDNW), Pre-estimate—Whitening—Post-estimate (PWP) loop, to reduce both photon and thermal noise in HSI. Previously, we proposed a method based on multidimensional wavelet packet transform and multi-way Wiener filter which performs both white noise and spectral dimensionality reduction, referred to as MWPT-MWF, which was restricted to white noise. We get inspired from this MWPT-MWF to develop a new iterative method for reducing photon and thermal noise. Firstly, the hyperspectral noise parameters estimation (HYNPE) algorithm is used to estimate the noise parameters, the SD noise is converted to an additive white Gaussian noise by pre-whitening procedure and then the whitened HSI is denoised by the proposed method SDNW-MWPT-MWF. As comparative experiments, the Multiple Linear Regression (MLR) based denoising method and tensor-based Multiway Wiener Filter (MWF) are also used in the denoising framework. An HSI captured by Reflective Optics System Imaging Spectrometer (ROSIS) is used in the experiments and the denoising performances are assessed from various aspects: the noise whitening performance, the Signal-to-Noise Ratio (SNR), and the classification performance. The results on the real-world airborne hyperspectral image HYDICE (Hyperspectral Digital Imagery Collection Experiment) are also presented and analyzed. These experiments show that it is worth taking into account noise signal-dependency hypothesis for processing HYDICE and ROSIS HSIs.

Hyperspectral Image Spatial Super-Resolution via 3D Full Convolutional Neural Network

Hyperspectral images are well-known for their fine spectral resolution to discriminate different materials. However, their spatial resolution is relatively low due to the trade-off in imaging sensor technologies, resulting in limitations in their applications. Inspired by recent achievements in convolutional neural network (CNN) based super-resolution (SR) for natural images, a novel three-dimensional full CNN (3D-FCNN) is constructed for spatial SR of hyperspectral images in this paper. Specifically, 3D convolution is used to exploit both the spatial context of neighboring pixels and spectral correlation of neighboring bands, such that spectral distortion when directly applying traditional CNN based SR algorithms to hyperspectral images in band-wise manners is alleviated. Furthermore, a sensor-specific mode is designed for the proposed 3D-FCNN such that none of the samples from the target scene are required for training. Fine-tuning by a small number of training samples from the target scene can further improve the performance of such a sensor-specific method. Extensive experimental results on four benchmark datasets from two well-known hyperspectral sensors, namely hyperspectral digital imagery collection experiment (HYDICE) and reflective optics system imaging spectrometer (ROSIS) sensors, demonstrate that our proposed 3D-FCNN outperforms several existing SR methods by ensuring higher quality both in reconstruction and spectral fidelity.

Hyperspectral Anomaly Detection Using Attribute Profiles

Over the past decade, anomaly detection has been an alluring topic in hyperspectral imagery. Despite most anomaly detection methods that focus on spectral information for detecting targets, this letter proposes a method in which both spectral and spatial information about the hyperspectral image (HSI) has been utilized for detecting anomalies. The differential attribute profile anomaly detection (DAPAD) method utilizes principal component analysis and DAP to extract spectral and spatial information from HSI, respectively. DAPs can model different kinds of structural information in a scene, which makes better extraction of spatial information. The proposed method is applied in experiments on Hyperspectral Digital Imagery Collection Experiment remote sensing data and the experimental results confirm the DAPAD method’s superiority over five commonly used state-of-the-art anomaly detection methods.

HS remote sensing image restoration using fusion with MS images by EM algorithm

Remote sensing images are widely used for different areas from mineral exploration to agricultural applications and poor quality of hyperspectral (HS) images will directly have adverse effect on these applications. In this study, a method is proposed to restore degraded HS images. To achieve this aim, another multispectral (MS) observation of the same scene is supposed to be available and restoration is fulfilled by fusion of HS images and MS images. The proposed method gains maximum a posteriori estimation and is based on expectation maximisation algorithm. Deblurring and denoising are performed separately. Deblurring is done in spatial domain via non-overlapping blocks, whereas denoising is implemented in wavelet domain. To represent the coefficients in wavelet domain, instead of multinormal model, Gaussian scale mixture is exploited. The proposed method is validated on airborne visible/infrared imaging spectrometer (AVIRIS) and HS digital imagery collection experiment (HYDICE) databases and experimental results signify that the proposed method outperforms state-of-the-art techniques cited in the literature and signal-to-noise ratio is improved as much as 15.71 dB for Moffett database and 16.26 dB for HYDICE database.

An Integrated Framework for the Spatio–Temporal–Spectral Fusion of Remote Sensing Images

Remote sensing satellite sensors feature a tradeoff between the spatial, temporal, and spectral resolutions. In this paper, we propose an integrated framework for the spatio–temporal–spectral fusion of remote sensing images. There are two main advantages of the proposed integrated fusion framework: it can accomplish different kinds of fusion tasks, such as multiview spatial fusion, spatio–spectral fusion, and spatio–temporal fusion, based on a single unified model, and it can achieve the integrated fusion of multisource observations to obtain high spatio–temporal–spectral resolution images, without limitations on the number of remote sensing sensors. The proposed integrated fusion framework was comprehensively tested and verified in a variety of image fusion experiments. In the experiments, a number of different remote sensing satellites were utilized, including IKONOS, the Enhanced Thematic Mapper Plus (ETM+), the Moderate Resolution Imaging Spectroradiometer (MODIS), the Hyperspectral Digital Imagery Collection Experiment (HYDICE), and Systeme Pour l' Observation de la Terre-5 (SPOT-5). The experimental results confirm the effectiveness of the proposed method.

Comparing support vector machine contextual approaches for urban area classification

ABSTRACTSupport vector machine (SVM) has been receiving a great deal of attention for remote sensing data classification. Although the original formulation of this method does not incorporate contextual information, lately different formulations have been proposed to incorporate such information, with the aim of improving the mapping accuracy. In general, these proposals modify the SVM training phase or integrate the SVM classifications in stochastic models. Recently, two new contextual versions of SVM, context adaptive and competitive translative SVM (CaSVM and CtSVM, respectively), were proposed in literature. In this work, two case studies of urban area classification, using IKONOS-II and hyperspectral digital imagery collection experiment (HYDICE) data sets were conducted to compare SVM, SVM integrated with the iterated conditional modes (ICM) stochastic algorithm, SVM smoothed using the mode filter and the recent approaches CaSVM and CtSVM. The results indicated that although it possesses a high computational cost, the CaSVM method was able to produce classification results with similar accuracy (using kappa coefficient) to those obtained using SVM integrated with ICM (SVM + ICM) and the mode filter (SVM + Mode), all of them found statistically superior to the SVM result at 95% confidence level for the IKONOS-II image. For HYDICE image, all results were found statistically insignificant at 95% confidence level. Investigation of what happens at transition regions between classes, however, showed that some methods can present superior performance. To this objective, a new performance measure, called upsilon coefficient, was introduced in this work, which measures the impact that the smoothing effect, typical of contextual methods, can have in distorting the edges between regions. With this new measure was found that CaSVM is the one which has better performance followed with SVM + ICM.

Discriminative spatial-spectral manifold embedding for hyperspectral image classification

In hyperspectral images (HSI) classification, it is important to combine multiple features of a certain pixel in both spatial and spectral domains to improve the classification accuracy. To achieve this goal, this article proposes a novel spatial-spectral feature dimensionality reduction algorithm based on manifold learning. For each feature, a graph Laplacian matrix is constructed based on discriminative information from training samples, and then the graph Laplacian matrices of the various features are linearly combined using a set of empirically defined weights. Finally, the feature mapping is obtained by an eigen-decomposition problem. Based on the classification results of the public Indiana Airborne Visible Infrared Imaging Spectrometer dataset and Texas Hyperspectral Digital Imagery Collection Experiment data set, the technical accuracies show that our method achieves superior performance compared to some representative HSI feature extraction and dimensionality reduction algorithms.

Principal Component Reconstruction Error for Hyperspectral Anomaly Detection

In this letter, a reliable, simple, and intuitive approach for hyperspectral imagery (HSI) anomaly detection (AD) is presented. This method, namely, the global iterative principal component analysis (PCA) reconstruction-error-based anomaly detector (GIPREBAD), examines AD by computing errors (residuals) associated with reconstructing the original image using PCA projections. PCA is a linear transformation and feature extraction process commonly used in HSI and frequently appears in operation prior to any AD task. PCA features represent a projection of the original data into lower-dimensional subspace. An iterative approach is used to mitigate outlier influence on background covariance estimates. GIPREBAD results are provided using receiver-operating-characteristic curves for HSI from the hyperspectral digital imagery collection experiment. Results are compared against the Reed-Xiaoli (RX) algorithm, the linear RX (LRX) algorithm, and the support vector data description (SVDD) algorithm. The results show that the proposed GIPREBAD method performs favorably compared with RX, LRX, and SVDD and is both intuitively and computationally simpler than either RX or SVDD.

Hyperspectral Image Resolution Enhancement Using High-Resolution Multispectral Image Based on Spectral Unmixing

In this paper, a hyperspectral (HS) image resolution enhancement algorithm based on spectral unmixing is proposed for the fusion of the high-spatial-resolution multispectral (MS) image and the low-spatial-resolution HS image (HSI). As a result, a high-spatial-resolution HSI is reconstructed based on the high spectral features of the HSI represented by endmembers and the high spatial features of the MS image represented by abundances. Since the number of endmembers extracted from the MS image cannot exceed the number of bands in least-squares-based spectral unmixing algorithm, large reconstruction errors will occur for the HSI, which degrades the fusion performance of the enhanced HSI. Therefore, in this paper, a novel fusion framework is also proposed by dividing the whole image into several subimages, based on which the performance of the proposed spectral-unmixing-based fusion algorithm can be further improved. Finally, experiments on the Hyperspectral Digital Imagery Collection Experiment and Airborne Visible/Infrared Imaging Spectrometer data demonstrate that the proposed fusion algorithms outperform other famous fusion techniques in both spatial and spectral domains.

Reduction of Signal-Dependent Noise From Hyperspectral Images for Target Detection

Tensor-decomposition-based methods for reducing random noise components in hyperspectral images (HSIs), both dependent and independent from signal, are proposed. In this paper, noise is described by a parametric model that accounts for the dependence of noise variance on the signal. This model is thus suitable for the cases where photon noise is dominant compared with the electronic noise contribution. To denoise HSIs distorted by both signal-dependent (SD) and signal-independent (SI) noise, some hybrid methods, which reduce noise by two steps according to the different statistical properties of those two types of noise, are proposed in this paper. The first one, named as the PARAFACSI- PARAFACSD method, uses a multilinear algebra model, i.e., parallel factor analysis (PARAFAC) decomposition, twice to remove SI and SD noise, respectively. The second one is a combination of the well-known multiple-linear-regression-based approach termed as the HYperspectral Noise Estimation (HYNE) method and PARAFAC decomposition, which is named as the HYNE-PARAFAC method. The last one combines the multidimensional Wiener filter (MWF) method and PARAFAC decomposition and is named as the MWF-PARAFAC method. For HSIs distorted by both SD and SI noise, first, most of the SI noise is removed from the original image by PARAFAC decomposition, the HYNE method, or the MWF method based on the statistical property of SI noise; then, the residual SD components can be further reduced by PARAFAC decomposition due to its own statistical property. The performances of the proposed methods are assessed on simulated HSIs. The results on the real-world airborne HSI Hyperspectral Digital Imagery Collection Experiment (HYDICE) are also presented and analyzed. These experiments show that it is worth taking into account noise signal-dependence hypothesis for processing HYDICE data.

An Unsupervised Spectral Matching Classifier Based on Artificial DNA Computing for Hyperspectral Remote Sensing Imagery

Hyperspectral remote sensing image clustering, with the large volume, high dimensions, and temporal-spatial spectral diversity, is a challenging task due to finding interesting clusters in the sparse feature space. In this paper, a novel hyperspectral clustering algorithm, namely, an unsupervised spectral matching classifier based on artificial DNA computing (UADSM), is proposed to perform the task of clustering different ground objects in specific spectral DNA feature encoding subspaces. UADSM builds up the clustering framework with the spectral encoding, optimizing, and matching mechanism by introducing the basic notions and operators of artificial DNA computing. By discretized spectral DNA feature encoding processing, the spectral shape, amplitude, and slope features of the hyperspectral data are extracted. Furthermore, the optimal clustering centers in the form of DNA strands can be found by recombining the DNA strands in the spectral DNA encoding subspace. Finally, a reasonable category for each spectral signature is automatically identified by the normalized spectral DNA similarity norm. The traditional clustering methods of k-means, ISODATA, fuzzy c-means classifier, and FCM and MoDEFC after principal component analysis transformation are provided to compare with the UADSM classifier, using Hyperspectral Digital Imagery Collection Experiment and Reflective Optics System Imaging Spectrometer hyperspectral images. The experimental results show that the UADSM classifier can achieve the best classification accuracy; hence, it is considered that the UADSM classifier is an effective clustering method for hyperspectral remote sensing imagery.

Dual force hybrid detector for hyperspectral images

This letter proposes a novel dual force hybrid detector for hyperspectral image (HSI) target detection. As a hybrid detector, we conduct the HSI spectral unmixing in a non-linear feature space mapped by locally linear embedding, and the image reconstruction error in the manifold embedded feature is considered as the primary force for detection. Then, we introduce the adaptive cosine/coherent estimate detector as the additional force. By combining the aforementioned two parts of detection results together, we obtain the final detection probability map. Based on the public data sets of (1) implanted metal detection in a hyperspectral digital imagery collection experiment image and (2) fabrics and vehicles detection in a Hyperspectral Mapper image, the technical results show that our method achieves superior performance compared to some state-of-the-art hyperspectral detectors.

An SVM Ensemble Approach Combining Spectral, Structural, and Semantic Features for the Classification of High-Resolution Remotely Sensed Imagery

In recent years, the resolution of remotely sensed imagery has become increasingly high in both the spectral and spatial domains, which simultaneously provides more plentiful spectral and spatial information. Accordingly, the accurate interpretation of high-resolution imagery depends on effective integration of the spectral, structural and semantic features contained in the images. In this paper, we propose a new multifeature model, aiming to construct a support vector machine (SVM) ensemble combining multiple spectral and spatial features at both pixel and object levels. The features employed in this study include a gray-level co-occurrence matrix, differential morphological profiles, and an urban complexity index. Subsequently, three algorithms are proposed to integrate the multifeature SVMs: certainty voting, probabilistic fusion, and an object-based semantic approach, respectively. The proposed algorithms are compared with other multifeature SVM methods including the vector stacking, feature selection, and composite kernels. Experiments are conducted on the hyperspectral digital imagery collection experiment DC Mall data set and two WorldView-2 data sets. It is found that the multifeature model with semantic-based postprocessing provides more accurate classification results (an accuracy improvement of 1-4% for the three experimental data sets) compared to the voting and probabilistic models.