Hyperspectral Imagery Processing Research Articles

Multiview Locally Linear Embedding for Spectral-Spatial Dimensionality Reduction of Hyperspectral Imagery

Dear Editor, Dimensionality reduction (DR) plays a prominent role in the processing of hyperspectral imagery. Considering the high dimensionality of multiple features, this letter presents a new unsupervised DR method named multiview locally linear embedding (MLLE), which captures the local linearity and global nonlinearity of the data sufficiently. We formulate MLLE as an optimization problem, where the diversity and complementarity of multiple features is fully exploited. An effective alternating optimization scheme is derived, and a linear model based on ridge regression is extended to alleviate the high correlation among single-view features. Experimental results on the Indian Pines and Pavia University datasets demonstrate the superiority of our proposed MLLE.

Unsupervised Recurrent Hyperspectral Imagery Super-Resolution Using Pixel-Aware Refinement

Unsupervised fusion-based hyperspectral imagery (HSI) super-resolution (SR) is an essential task of HSI processing, which aims to reconstruct a high-resolution (HR) HSI using only an observed low-resolution HSI and a conventional HR image. Although a large number of unsupervised HSI SR methods have been proposed, the heuristic handcrafted image priors adopted by the majority of these methods restrict their capacity to capture specific characteristics of the HSI, as well as their ability to generalize to noisy observation images. In this study, we investigate a fusion-based HSI SR framework with the deep image prior, in which the deep neural network (rather than a heuristic handcrafted image prior) is exploited to capture plenty of image statistics. Within this framework, we further propose an unsupervised recurrence-based HSI SR method using pixel-aware refinement, which utilizes the intermediate reconstruction results to self-supervise unsupervised learning. Due to containing the information of the image-specific characteristic, the proposed method achieves better performance, in terms of both accuracy and robustness to noise, compared with the existing methods. Extensive experiments on four HSI data sets demonstrate the effectiveness of the proposed method.

Underwater Hyperspectral Target Detection with Band Selection

Compared to multi-spectral imagery, hyperspectral imagery has very high spectral resolution with abundant spectral information. In underwater target detection, hyperspectral technology can be advantageous in the sense of a poor underwater imaging environment, complex background, or protective mechanism of aquatic organisms. Due to high data redundancy, slow imaging speed, and long processing of hyperspectral imagery, a direct use of hyperspectral images in detecting targets cannot meet the needs of rapid detection of underwater targets. To resolve this issue, a fast, hyperspectral underwater target detection approach using band selection (BS) is proposed. It first develops a constrained-target optimal index factor (OIF) band selection (CTOIFBS) to select a band subset with spectral wavelengths specifically responding to the targets of interest. Then, an underwater spectral imaging system integrated with the best-selected band subset is constructed for underwater target image acquisition. Finally, a constrained energy minimization (CEM) target detection algorithm is used to detect the desired underwater targets. Experimental results demonstrate that the band subset selected by CTOIFBS is more effective in detecting underwater targets compared to the other three existing BS methods, uniform band selection (UBS), minimum variance band priority (MinV-BP), and minimum variance band priority with OIF (MinV-BP-OIF). In addition, the results also show that the acquisition and detection speed of the designed underwater spectral acquisition system using CTOIFBS can be significantly improved over the original underwater hyperspectral image system without BS.

Validation of Information Products of Airborne Hyperspectral Imagery Processing

In regards to the problem of airborne hyperspectral imagery processing, most attention is paid to the construction of an automated classifier which implements computational procedures of recognizing forest vegetation of different species composition and age. The validation of the information products obtained as a result of processing the data of hyperspectral remote sensing is based on finding a maximum of a posteriori probability for selected classes of objects using the contemporary views of Markov random fields allowing a priori probabilities for each element of hyperspectral apparatus resolution to be calculated. The necessity of taking into account the noise component of the apparatus to enhance the accuracy of solving the applied problems under study is shown. Examples of comparing the data of hyperspectral imagery processing with the data of ground-based forest inventory for a selected territory are demonstrated.

A Probabilistic Hyperspectral Imagery Restoration Method

Hyperspectral image (HSI) restoration is an important task of hyperspectral imagery processing, which aims to improve the performance of the subsequent HSI interpretation and applications. Considering HSI is always influenced by multiple factors—such as Gaussian noise, stripes, dead pixels, etc.—we propose an HSI-oriented probabilistic low-rank restoration method to address this problem. Specifically, we treat the expected clean HSI as a low-rank matrix. We assume the distribution of complex noise obeys a mixture of Gaussian distributions. Then, the HSI restoration problem is casted into solving the clean HSI from its counterpart with complex noise. In addition, considering the rank number need to be assigned manually for existing low-rank based HSI restoration method, we propose to automatically determine the rank number of the low-rank matrix by taking advantage of hyperspectral unmixing. Experimental results demonstrate HSI image can be well restored with the proposed method.

Hyperspectral Unmixing via Total Variation Regularized Nonnegative Tensor Factorization

Hyperspectral unmixing decomposes a hyperspectral imagery (HSI) into a number of constituent materials and associated proportions. Recently, nonnegative tensor factorization (NTF)-based methods have been proposed for hyperspectral unmixing thanks to their capability in representing an HSI without any information loss. However, tensor factorization-based HSI processing approaches often suffer from low-signal-to-noise ratio condition of HSI and nonuniqueness of the solution. This problem can be effectively alleviated by introducing various spatial constraints into tensor factorization to suppress the noise and decrease the number of extreme, stationary, and saddle points. On the other hand, total variation (TV) adaptively promotes piecewise smoothness while preserving edges. In this paper, we propose a TV regularized matrix–vector NTF method. It takes advantage of tensor factorization in preserving global spectral–spatial information and the merits of TV in exploiting local spatial information, thus generating smooth abundance maps with preserved edges. Experimental results on synthetic and real-world data show that the proposed method outperforms the state-of-the-art methods.

Hyperspectral Remote Sensing Imagery Processing Focused on Forest Applications

Imaging spectrometers with hundreds of spectral channels in visible and infrared regions are designed by various companies to enhance the information content of the relevant hyperspectral imagery processing compared to common-used multispectral systems. We review some sources on this particular subject to show the priorities of the hyperspectral approach before the multispectral one in forest and agriculture applications. There is also a discussion about some results of the information products obtained by an imaging spectrometer produced in Russia for a test area, where the ground-based forest inventory map is available to compare the traditional approaches and the newly defined ones. The related applications concern the pattern recognition of forest classes with different species and age on the test area using the airborne hyperspectral imagery processing.

Initial study of Schroedinger eigenmaps for spectral target detection

Spectral target detection refers to the process of searching for a specific material with a known spectrum over a large area containing materials with different spectral signatures. Traditional target detection methods in hyperspectral imagery (HSI) require assuming the data fit some statistical or geometric models and based on the model, to estimate parameters for defining a hypothesis test, where one class (i.e., target class) is chosen over the other classes (i.e., background class). Nonlinear manifold learning methods such as Laplacian eigenmaps (LE) have extensively shown their potential use in HSI processing, specifically in classification or segmentation. Recently, Schroedinger eigenmaps (SE), which is built upon LE, has been introduced as a semisupervised classification method. In SE, the former Laplacian operator is replaced by the Schroedinger operator. The Schroedinger operator includes by definition, a potential term V that steers the transformation in certain directions improving the separability between classes. In this regard, we propose a methodology for target detection that is not based on the traditional schemes and that does not need the estimation of statistical or geometric parameters. This method is based on SE, where the potential term V is taken into consideration to include the prior knowledge about the target class and use it to steer the transformation in directions where the target location in the new space is known and the separability between target and background is augmented. An initial study of how SE can be used in a target detection scheme for HSI is shown here. In-scene pixel and spectral signature detection approaches are presented. The HSI data used comprise various target panels for testing simultaneous detection of multiple objects with different complexities.

Testing different classification methods in airborne hyperspectral imagery processing.

To enhance the efficiency of machine-learning algorithms of optical remote sensing imagery processing, optimization techniques are evolved of the land surface objects pattern recognition. Different methods of supervised classification are considered for these purposes, including the metrical classifier operating with Euclidean distance between any points of the multi-dimensional feature space given by registered spectra, the K-nearest neighbors classifier based on a majority vote for neighboring pixels of the recognized objects, the Bayesian classifier of statistical decision making, the Support Vector Machine classifier dealing with stable solutions of the mini-max optimization problem and their different modifications. We describe the related techniques applied for selected test regions to compare the listed classifiers.

Bayesian classifier applications of airborne hyperspectral imagery processing for forested areas

Pattern recognition problem is outlined in the context of textural and spectral analysis of remote sensing imagery processing. Main attention is paid to Bayesian classifier that can be used to realize the processing procedures based on parallel machine-learning algorithms and high-productive computers. We consider the maximum of the posterior probability principle and the formalism of Markov random fields for the neighborhood description of the pixels for the related classes of objects with the emphasis on forests of different species and ages. The energy category of the selected classes serves to account for the likelihood measure between the registered radiances and the theoretical distribution functions approximating remotely sensed data. Optimization procedures are undertaken to solve the pattern recognition problem of the texture description for the forest classes together with finding thin nuances of their spectral distribution in the feature space. As a result, possible redundancy of the channels for imaging spectrometer due to their correlations is removed. Difficulties are revealed due to different sampling data while separating pixels, which characterize the sunlit tops, shaded space and intermediate cases of the Sun illumination conditions on the hyperspectral images. Such separation of pixels for the forest classes is maintained to enhance the recognition accuracy, but learning ensembles of data need to be agreed for these categories of pixels. We present some results of the Bayesian classifier applicability for recognizing airborne hyperspectral images using the relevant improvements in separating such pixels for the forest classes on a test area of the 4×10km size encompassed by 13 airborne tracks, each forming the images by 500 pixels across the track and from 10,000 to 14,000 pixels along the track. The spatial resolution of each image is near to 1m from the altitude near to 2km above the ground level. The results of the hyperspectral imagery processing have shown that for the ensembles of data corresponding to the prevailing pine species and sunlit tree’s tops the young forests (13–26years old) and mature forests (106–136years old) on the test area are recognized with high accuracy, but difficulties are essential for intermediate ages (36–96years old) though some of these ages (for example, 47 and 76years old) are also recognized with high accuracy. This is due to insufficient sampling of the pixels relating to the sunlit tops for the intermediate ages using the classifier employed. A conclusion is made that a fusion of the passive imaging spectrometer and the active lidar (laser scanner) on the same gyro-stabilized airborne platform is needed to overcome these difficulties.

Thematic classification of hyperspectral images using conjugacy indicator

We consider an algorithm of hyperspectral images thematic classification using conjugacy indicator as a proximity measure. This measure is a generalized spectral angle mapper (SAM) implemented in hyperspectral imagery processing software ENVI. In this case, we use the cosine of an angle between considered vector and subspace, which is spanned by class vectors, instead of the cosine of an angle between considered vector and the mean vector of the class. Paper describes modification of a method based on partitioning of the class into subclasses and based on reduction of vectors to zero mean value. The results of synthetic experiments show higher classification quality than SAM.

Spectral Unmixing Model Based on Least Squares Support Vector Machine With Unmixing Residue Constraints

Spectral unmixing has been an important technique for hyperspectral imagery processing. In traditional spectral unmixing methods that are based on the linear spectral mixture model (LSMM), unmixing accuracy is limited by the inherent deficiency of the model. It was shown that the support vector machine (SVM) can be extended for spectral unmixing, based on the advantage that the SVM model can accommodate the variations within a relative pure class by using multiple pure samples instead of a single endmember for one class. In the SVM model, class label errors are considered in constraints. However, the errors concerned in spectral unmixing are the unmixing residue instead of the class label ones. This letter presents a method of imposing unmixing residue constraints on the least squares SVM unmixing model. The related problems, including deducing the closed-form solution and substituting the single endmember for multiple ones, were studied together. Experiments showed that the new SVM model was superior to the original SVM as well as the traditional LSMM in terms of unmixing residue, fractional abundance, and confused matrix criterions.

AutoGAD: An Improved ICA-Based Hyperspectral Anomaly Detection Algorithm

This work extends the emerging field of hyperspectral imagery (HSI) anomaly detectors and employs independent component (IC) analysis (ICA) to identify anomalous pixels. Using new techniques to fully automate feature extraction, feature selection, and anomaly pixel identification, an Autonomous Global Anomaly Detector has been developed for employment in an operational environment for real-time processing of HSI. Dimensionality reduction, which is the initial feature extraction prior to ICA, is effected through a geometric solution that estimates the number of retained principal components. The solution is based on the theory of the shape of the eigenvalue curve of the covariance matrix of spectral data containing noise. This research presents two new features, namely, potential anomaly signal-to-noise ratio and maximum pixel score, both of which are computed for each of the ICs to create a new 2-D feature space where the overlap between anomaly and nonanomaly classes is reduced. After anomaly feature selection, adaptive noise filtering is applied iteratively to suppress the background. Finally, a zero-detection histogram method is applied to the smoothed signals to identify anomaly locations to the user. After the algorithm is fully developed, a set of designed experiments are conducted to identify a reasonable set of input parameters.

Remote sensing of soils and vegetation: quantitative parameters retrieval using pattern-recognition techniques and forest stand structure assessment

New findings in numerical mathematics application are outlined in the context of remote-sensing science (RSS), which is an interdisciplinary concept encompassing the proposed modelling approach, multi-spectral (the number of spectral bands is usually not more than 10) and hyper-spectral (hundreds of spectral bands) airspace data processing, validation campaigns of ground-based data use. The modelling approach involves the direct problem of the outgoing radiation calculation and the inverse problem of the retrieval of land surface parameters. Pattern-recognition procedures use a mathematical technique of imagery classification combined with retrieval techniques for each processed pixel. Validation experiments are conducted to compile surface flux data, biogeochemical samples, measurements from specially equipped towers and other parts of data collection for selected plots. The following RSS key areas are considered: the two-dimensional vegetation structure description, instead of the commonly used leaf area index (LAI) improvements; the green phytomass (for leaves/needles) assessment as an alternative to the ‘indices concept’ (Normalized Difference Vegetation Index (NDVI), etc.) using multi-spectral and hyper-spectral imagery processing. Results are shown as the net primary productivity (NPP) assessment for the selected transect of the southern biome of boreal forest in Russia using airborne remote-sensing data, ground-based forest taxation results and validation techniques of pixel-by-pixel classification together with the forest phytomass retrievals that serve to present the relevant results of remote-sensing data processing in terms of the NPP values.

Fast real-time onboard processing of hyperspectral imagery for detection and classification

Remotely sensed hyperspectral imagery has many important applications since its high-spectral resolution enables more accurate object detection and classification. To support immediate decision-making in critical circumstances, real-time onboard implementation is greatly desired. This paper investigates real-time implementation of several popular detection and classification algorithms for image data with different formats. An effective approach to speeding up real-time implementation is proposed by using a small portion of pixels in the evaluation of data statistics. An empirical rule of an appropriate percentage of pixels to be used is investigated, which results in reduced computational complexity and simplified hardware implementation. An overall system architecture is also provided.

Field Programmable Gate Arrays (FPGA) for Pixel Purity Index Using Blocks of Skewers for Endmember Extraction in Hyperspectral Imagery

Pixel purity index (PPI) is a popular endmember extraction algorithm because of its availability in the ENVI software. Recently, its hardware implementation has been explored by the so-called block of skewers (BOS) that can be used to generate desired skewers required by the PPI with significantly reduced computational complexity. This paper investigates the feasibility of field programmable gate arrays (FPGAs) design for the BOS of a modified PPI (called automatic PPI), referred to as BOS-APPI in high speed computation. Several block designs based on cube, pyramid, square and tetrahedron are also developed for the BOS-APPI along with their FPGA implementations as new reconfigurable systems to perform fast and onboard processing of hyperspectral imagery. Finally, the proposed FPGA designs are realized and substantiated by simulations.

Commodity cluster-based parallel processing of hyperspectral imagery

The rapid development of space and computer technologies has made possible to store a large amount of remotely sensed image data, collected from heterogeneous sources. In particular, NASA is continuously gathering imagery data with hyperspectral Earth observing sensors such as the Airborne Visible-Infrared Imaging Spectrometer (AVIRIS) or the Hyperion imager aboard Earth Observing-1 (EO-1) spacecraft. The development of fast techniques for transforming the massive amount of collected data into scientific understanding is critical for space-based Earth science and planetary exploration. This paper describes commodity cluster-based parallel data analysis strategies for hyperspectral imagery, a new class of image data that comprises hundreds of spectral bands at different wavelength channels for the same area on the surface of the Earth. An unsupervised technique that integrates the spatial and spectral information in the image data using multi-channel morphological transformations is parallelized and compared to other available parallel algorithms. The code's portability, reusability and scalability are illustrated by using two high-performance parallel computing architectures: a distributed memory, multiple instruction multiple data (MIMD)-style multicomputer at European Center for Parallelism of Barcelona, and a Beowulf cluster at NASA's Goddard Space Flight Center. Experimental results suggest that Beowulf clusters are a source of computational power that is both accessible and applicable to obtaining results in valid response times in information extraction applications from hyperspectral imagery.

Accuracy assessment for detection of leafy spurge with hyperspectral imagery

When flowering, leafy spurge (Euphorbia esula L.) has conspicuous yellow-green bracts that are spectrally distinct from other vegetation and may be distinguished with hyperspectral remote sensing. In July 1999, Airborne Visible Infrared Imaging Spectrometer (AVIRIS) data were acquired in northeastern Wyoming, near Devils Tower National Monument. Using the reflectance spectrum of flowering leafy spurge, leafy spurge occurrence was determined using a new method of spectral mixture analysis, Mixture Tuned Matched Filtering (MTMF). Ground reference data (146 sites) were obtained 2 weeks before and after AVIRIS overflight to test the classification accuracy of leafy spurge. For 3 land cover types: mixed prairie, riparian, and coniferous woodlands, the presence or absence leafy spurge was detected with an overall accuracy of 95% using a 0.10 threshold for detection. Differences in classification thresholds resulted in a trade-off between false positives, pixels that were mapped as leafy spurge but did not contain leafy spurge on the ground, and false negatives, areas that had leafy spurge on the ground but were not mapped as leafy spurge. Detection of leafy spurge occurrence was best for mixed prairie and riparian cover types, and somewhat less successful for conifer woodlands because of interference from tree crowns and their shadows. The advantage of the MTMF technique is it allows automated processing of hyperspectral imagery to generate accurate maps of leafy spurge occurrence.

Accuracy assessment for detection of leafy spurge with hyperspectral imagery

When flowering, leafy spurge (Euphorbia esula L.) has conspicuous yellow-green bracts that are spectrally distinct from other vegetation and may be distinguished with hyperspectral remote sensing. In July 1999, Airborne Visible Infrared Imaging Spectrometer (AVIRIS) data were acquired in northeastern Wyoming, near Devils Tower National Monument. Using the reflectance spectrum of flowering leafy spurge, leafy spurge occurrence was determined using a new method of spectral mixture analysis, Mixture Tuned Matched Filtering (MTMF). Ground reference data (146 sites) were obtained 2 weeks before and after AVIRIS overflight to test the classification accuracy of leafy spurge. For 3 land cover types: mixed prairie, riparian, and coniferous woodlands, the presence or absence leafy spurge was detected with an overall accuracy of 95% using a 0.10 threshold for detection. Differences in classification thresholds resulted in a trade-off between false positives, pixels that were mapped as leafy spurge but did not contain leafy spurge on the ground, and false negatives, areas that had leafy spurge on the ground but were not mapped as leafy spurge. Detection of leafy spurge occurrence was best for mixed prairie and riparian cover types, and somewhat less successful for conifer woodlands because of interference from tree crowns and their shadows. The advantage of the MTMF technique is it allows automated processing of hyperspectral imagery to generate accurate maps of leafy spurge occurrence.

Preliminary investigation of submerged aquatic vegetation mapping using hyperspectral remote sensing.

The use of airborne hyperspectral remote sensing imagery for automated mapping of submerged aquatic vegetation (SAV) in the tidal Potomac River was investigated for near to real-time resource assessment and monitoring. Airborne hyperspectral imagery and field spectrometer measurements were obtained in October of 2000. A spectral library database containing selected ground-based and airborne sensor spectra was developed for use in image processing. The spectral library is used to automate the processing of hyperspectral imagery for potential real-time material identification and mapping. Field based spectra were compared to the airborne imagery using the database to identify and map two species of SAV (Myriophyllum spicatum and Vallisneria americana). Overall accuracy of the vegetation maps derived from hyperspectral imagery was determined by comparison to a product that combined aerial photography and field based sampling at the end of the SAV growing season. The algorithms and databases developed in this study will be useful with the current and forthcoming space-based hyperspectral remote sensing systems.