Spectral Endmember Extraction Research Articles

Quantitative Mineral Mapping of Drill Core Surfaces II: Long-Wave Infrared Mineral Characterization UsingμXRF and Machine Learning

AbstractLong-wave infrared (LWIR) spectra can be interpreted using a Random Forest machine learning approach to predict mineral species and abundances. In this study, hydrothermally altered carbonate rock core samples from the Fourmile Carlin-type Au discovery, Nevada, were analyzed by LWIR and micro-X-ray fluorescence (μXRF). Linear programming-derived mineral abundances from quantified μXRF data were used as training data to construct a series of Random Forest regression models. The LWIR Random Forest models produced mineral proportion estimates with root mean square errors of 1.17 to 6.75% (model predictions) and 1.06 to 6.19% (compared to quantitative X-ray diffraction data) for calcite, dolomite, kaolinite, white mica, phlogopite, K-feldspar, and quartz. These results are comparable to the error of proportion estimates from linear spectral deconvolution (±7–15%), a commonly used spectral unmixing technique. Having a mineralogical and chemical training data set makes it possible to identify and quantify mineralogy and provides a more robust and meaningful LWIR spectral interpretation than current methods of utilizing a spectral library or spectral end-member extraction. Using the method presented here, LWIR spectroscopy can be used to overcome the limitations inherent with the use of short-wave infrared (SWIR) in fine-grained, low reflectance rocks. This new approach can be applied to any deposit type, improving the accuracy and speed of infrared data interpretation.

Parallel Implementation of Spatial–Spectral Endmember Extraction on Graphic Processing Units

The identification of pure spectral signatures (endmembers) in remotely sensed hyperspectral images has traditionally focused on the spectral information alone. Recently, techniques such as the spatial–spectral endmember extraction (SSEE) have incorporated both the spectral and the spatial information contained in the scene. Since hyperspectral images contain very detailed information in the spatial and spectral domain, the integration of these two sources of information generally comes with a significant increase in computational complexity. In this paper, we develop a new computationally efficient implementation of SSEE using commodity graphics processing units (GPUs). The relevance of GPUs comes from their very low cost, compact size, and the possibility to obtain significant acceleration factors by exploiting properly the GPU hardware architecture. Our experimental results, focused on evaluating the candidate endmembers produced by SSEE and also the computational performance of the GPU implementation, indicated significant acceleration factors that allow exploiting the SSEE method in computationally efficient fashion.

Mapping of NiCu–PGE ore hosting ultramafic rocks using airborne and simulated EnMAP hyperspectral imagery, Nunavik, Canada

This study first investigates using AISA airborne hyperspectral imagery (2m spatial resolution) to produce detailed lithologic maps in a subarctic region (Nunavik, Canada) where ultramafic rock units associated with Ni–Cu-(PGE) mineralization are exposed in the presence of lichen coatings. Twenty AISA flight-lines were radiometrically leveled and merged to form a 10×20km mosaic, which then served to generate a simulated spaceborne EnMAP scene (30m spatial resolution) using the End-to-End Simulation Tool to assess the sensors mapping capabilities. Spatial Spectral Endmember Extraction was used to derive spectral endmembers for the AISA and EnMAP data. Image endmembers were compared with spectral measurements of field samples to assess how well key rock types were represented. Results show that the AISA imagery provided a better representation of mafic and ultramafic rock types compared with the EnMAP simulation. Endmembers were then used to map the distribution of geological materials using Iterative Spectral Mixture Analysis. Results indicate the airborne data provided more detailed maps compared with EnMAP simulated data. However, EnMAP data could still discriminate and map broad scale lithological units, specifically mafic and ultramafic rocks. This study demonstrates the feasibility of EnMAP to provide large scale reconnaissance mapping capability of geologic materials over vast subarctic and arctic regions (potentially 30×5000km of imagery per day) using expert knowledge combined with automated spectral analysis methods.

Using airborne hyperspectral data to characterize the surface pH and mineralogy of pyrite mine tailings

Acid mine drainage (AMD) is a key concern of the mining industry due to its impact on the quality of water and soils surrounding mine waste deposits. Acid mine drainage derives from the oxidation of metal sulphides, e.g. pyrite (FeS2), exposed to oxygen and water. The leachate acidity is capable of releasing heavy metals contained in the mining waste rock, which can affect water quality and lead to metal enrichment in sediments and potentially resulting in ecosystem degradation. Predicting tailings leachate pH is key to the management of sulfide-bearing mine wastes and is an emerging remote sensing application with limited studies having been realized. Such a capability would supplement traditional methods (i.e. ground surveys) that are challenging to implement due to the extent and large volume of mine waste.This study reports regional scale tailings mineral maps generated from airborne hyperspectral information of the Sotiel-Migollas complex in Spain and pinpoints sources of AMD. The extraction of spectral endmembers from imagery revealed twenty six endmembers for tailings material that represent mostly mineral mixtures. From these, eleven spectral groups were defined, each encompassing minor variations in mineral mixtures. The mineral maps resulting from the use of these endmembers for the detailed investigation of four tailings serve as indicators of the metal, sulphate, and pH levels of the AMD solution at the time of mineral precipitation. Predicted mineralogy was assessed using spectra from samples collected in the field and associated X-ray diffraction measurements.We also discuss the relative merits of the minerals maps of this study and soil leachate pH maps that we previously reported for the same locality using the same airborne data. The pH maps tend to provide predictions consistent with the mineralogy predicted from the mineral maps and the field and laboratory evidence. The pH maps offer information on the pH conditions of the tailings thus giving an insight on the different types of oxidation reactions that may occur.

Spatial Sub-Sampling Using Local Endmembers for Adapting OSP and SSEE for Large-Scale Hyperspectral Surveys

Airborne and satellite hyperspectral sensors can collect significant quantities of data, commonly terabits in size. Thus, there is a need for the development of computational cost effective algorithms that speed up processing, yet retain essential data quality and information. Endmember extraction is critical in the processing chain of hyperspectral data. Existing methods have primarily been demonstrated on small data sets so their usefulness with large data sets has not been fully explored. The objective of this paper is to adapt the Orthogonal Subspace Projection (OSP) and Spatial Spectral Endmember Extraction (SSEE) algorithms, such that they run efficiently on large data sets without the loss of global image endmember quality. This is demonstrated using a simulated and two real hyperspectral images, the last of which is a multi- flight-line survey. The two adapted methods (OSP SS and SSEE SS) make use of spatial sub-sampling via local endmember extraction to reduce the size of the original data set. This paper will demonstrate that this type of spatial sub-sampling retains the full volume of the data, and thus, the vertices of the simplex that represent the global image endmembers. The results also show that computational time is reduce by more than half using the adapted methods. For the large multi- flight-line survey SSEE SS was able to better retain endmembers of natural materials compared with OSP SS. These results illustrate the potential of spatial sub-sampling for other endmember extraction tools when dealing with multi-flight-line surveys.

Improving Spatial–Spectral Endmember Extraction in the Presence of Anomalous Ground Objects

Endmember extraction (EE) has been widely utilized to extract spectrally unique and singular spectral signatures for spectral mixture analysis of hyperspectral images. Recently, spatial–spectral EE (SSEE) algorithms have been proposed to achieve superior performance over spectral EE (SEE) algorithms by taking both spectral similarity and spatial context into account. However, these algorithms tend to neglect anomalous endmembers that are also of interest. Therefore, in this paper, an improved SSEE (iSSEE) algorithm is proposed to address such limitation of conventional SSEE algorithms by accounting for both anomalous and normal endmembers. By developing simplex projection and simplex complementary projection, all the hyperspectral pixels are projected into a simplex determined by the normal endmembers extracted in conventional SSEE algorithms. As a result, anomalous endmembers are identified iteratively by utilizing the <formula formulatype="inline" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex Notation="TeX">$l_{2}^{\infty}$</tex></formula> norm to find the maximum simplex complementary projection. In order to determine how many anomalous endmembers are to be extracted, a novel Residual-be-Noise Probability-based algorithm is also proposed by elegantly utilizing the spatial-purity map generated in the previous SSEE step. Experimental results on both synthetic and real datasets demonstrate that simplex projection errors can be significantly reduced by identifying both anomalous and normal endmembers in the proposed iSSEE algorithm. It is also confirmed that the performance of the proposed iSSEE algorithm clearly outperforms that of SEE algorithms since both spatial context and spectral similarity are utilized.

Region-Based Spatial Preprocessing for Endmember Extraction and Spectral Unmixing

Linear spectral unmixing is an important task in remotely sensed hyperspectral data exploitation. This approach first identifies a collection of spectrally pure constituent spectra, called endmembers, and then expresses the measured spectrum of each mixed pixel as a combination of endmembers weighted by fractions or abundances that indicate the proportion of each endmember in the pixel. Over the last decade, several algorithms have been developed for automatic extraction of spectral endmembers from hyperspectral data, with many of them relying exclusively on the spectral information. In this letter, we develop a novel unsupervised spatial preprocessing (SPP) module which adopts a region-based approach for the characterization of each endmember class prior to endmember identification using spectral information. The proposed approach can be combined with any spectral-based endmember extraction technique. Our method is validated using both synthetic scenes constructed using fractals and a real hyperspectral data set collected by NASA's Airborne Visible Infrared Imaging Spectrometer over the Cuprite Mining District in Nevada and further compared with previous efforts in the same direction such as the spatial-spectral endmember extraction, automatic morphological endmember extraction, or SPP methods.

Impact of JPEG2000 compression on endmember extraction and unmixing of remotely sensed hyperspectral data

Lossy hyperspectral image compression has received considerable interest in recent years due to the extremely high dimensionality of the data. However, the impact of lossy compression on spectral unmixing techniques has not been widely studied. These techniques characterize mixed pixels (resulting from insufficient spatial resolution) in terms of a suitable combination of spectrally pure substances (called endmembers) weighted by their estimated fractional abundances. This paper focuses on the impact of JPEG2000-based lossy compression of hyperspectral images on the quality of the endmembers extracted by different algorithms. The three considered algorithms are the orthogonal subspace projection (OSP), which uses only spatial information, and the automatic morphological endmember extraction (AMEE) and spatial spectral endmember extraction (SSEE), which integrate both spatial and spectral information in the search for endmembers. The impact of compression on the resulting abundance estimation based on the endmembers derived by different methods is also substantiated. Experimental results are conducted using a hyperspectral data set collected by NASA Jet Propulsion Laboratory over the Cuprite mining district in Nevada. The experimental results are quantitatively analyzed using reference information available from U.S. Geological Survey, resulting in recommendations to specialists interested in applying endmember extraction and unmixing algorithms to compressed hyperspectral data.

Integration of spatial–spectral information for the improved extraction of endmembers

Spectral-based image endmember extraction methods hinge on the ability to discriminate between pixels based on spectral characteristics alone. Endmembers with distinct spectral features (high spectral contrast) are easy to select, whereas those with minimal unique spectral information (low spectral contrast) are more problematic. Spectral contrast, however, is dependent on the endmember assemblage, such that as the assemblage changes so does the “relative” spectral contrast of each endmember to all other endmembers. It is then possible for an endmember to have low spectral contrast with respect to the full image, but have high spectral contrast within a subset of the image. The spatial–spectral endmember extraction tool (SSEE) works by analyzing a scene in parts (subsets), such that we increase the spectral contrast of low contrast endmembers, thus improving the potential for these endmembers to be selected. The SSEE method comprises three main steps: 1) application of singular value decomposition (SVD) to determine a set of basis vectors that describe most of the spectral variance for subsets of the image; 2) projection of the full image data set onto the locally defined basis vectors to determine a set of candidate endmember pixels; and, 3) imposing spatial constraints for averaging spectrally similar endmembers, allowing for separation of endmembers that are spectrally similar, but spatially independent. The SSEE method is applied to two real hyperspectral data sets to demonstrate the effects of imposing spatial constraints on the selection of endmembers. The results show that the SSEE method is an effective approach to extracting image endmembers. Specific improvements include the extraction of physically meaningful, low contrast endmembers that occupy unique image regions.

Parallel Implementation of Endmember Extraction Algorithms From Hyperspectral Data

Automated extraction of spectral endmembers is a crucial task in hyperspectral data analysis. In most cases, the computational complexity of endmember extraction algorithms is very high, in particular, for very high-dimensional datasets. However, the intrinsic properties of available techniques are amenable to the design of parallel implementations. In this letter, we evaluate several parallel algorithms that represent three representative approaches to the problem of extracting endmembers. Two parallel algorithms have been selected to represent a first class of algorithms based on convex geometry concepts. In particular, we develop parallel implementations of approximate versions of the N-FINDR and pixel purity index algorithms, along with a parallel hybrid of both techniques. A second class is given by algorithms based on constrained error minimization and represented by a parallel version of the iterative error analysis algorithm. Finally, a parallel version of the automated morphological endmember extraction algorithm is also presented and discussed. This algorithm integrates the spatial and spectral information as opposed to the other discussed algorithms, a feature that introduces additional considerations for its parallelization. The proposed algorithms are quantitatively compared and assessed in terms of both endmember extraction accuracy and parallel efficiency, using standard AVIRIS hyperspectral datasets. Performance data are measured on Thunderhead, a parallel supercomputer at NASA's Goddard Space Flight Center

Diagnostic classification of urothelial cells in urine cytology specimens using exclusively spectral information.

Although cytologic evaluation of urine specimens is a standard procedure in the diagnosis and follow-up of bladder carcinoma, its sensitivity and specificity are low. Cytopathologic diagnoses are driven primarily by spatial relations or morphology. Although color enhances the pathologist's perception of the specimen, spectral information plays a minimal role in diagnostic processes. Recently, methods have been developed to capture and analyze spectral information from clinical specimens. In the current study, the authors determined the classification value of spectral information by testing its ability to discriminate between malignant and benign urothelial cells in cytology specimens. Multiple images of benign urothelial cells (n = 39) and urothelial carcinoma cells (n = 35) were collected at serial wavelengths using a liquid crystal tunable optical filter and composited into a mosaic using ENVI (Environment for Visualizing Images) software. Through minimum noise fractionation and principal component analysis, the spectral information in the mosaic was compressed into a 29-dimensional scatter plot. The data generated were analyzed using visual and spectral end member extraction on both the original data set and a second independent data set (test set). One area of spectral clustering in the scatter plot segmented with carcinoma cells exclusively (100% specific), but was not present in every cell (approximately 50%), which may indicate that these spectral profiles are present in a subpopulation of malignant cells or at specific points of their cell cycle. Using ENVI algorithms, the authors found that a particular classification spectrum (end member 9) and its closest relatives identified malignant cell clusters, with a sensitivity and specificity that reached 82% and 81%, respectively. To validate this mechanism in a test set, a second mosaic comprised of 15 benign and 15 malignant clusters was analyzed using end member 9, resulting in a combined sensitivity and specificity of 73%. The results of the current study demonstrate that spectral information, in the complete absence of morphologic or spatial information, allows discrimination of benign and malignant urothelial cells in routine urine cytology specimens. The authors believe that this novel technology, combined with spatial analysis, has the potential to serve as an ancillary test for improved detection of bladder carcinoma.

ANALYSIS OF LAND SAT-5 TM IMAGERY FOR EXTRACTING AQUACULTURE FARMS INFORMATION IN THE KOREAN COASTAL WATERS

The objective of the present study is to compare certain conventional satellite image processing techniques with the recently evolved linear spectral unmixing method and to ascertain the best possible technique that can identify and position of aquaculture farms accurately in and around the Younggwang coastal region of Korea. Various conventional techniques existed to extract such information are spectral enhancement and classification. However, these techniques performed on the Landsat-TM imagery do not seem to yield accurate information about the aquaculture farms, and instead lead to misinterpretation and misclassification. A large discrepancy between the actual and extracted information results from spectral confusion and inadequate spatial resolution of the sensor, which leads to occurrence of mixture pixels or mixels, which are known to be the source of errors in the classified image. To over come this problem, more recently evolved methods such as step-by-step iterative partial spectral end-member extraction linear spectral unmixing methods are attempted. Large errors in extraction of aquaculture farms information through the conventional classification techniques are significantly minimized with the step-by-step iterative partial spectral end-member extraction approach and the accuracy of classification is further improved with linear spectral unmixing approach. The aquaculture fraction derived from unmxing of TM image data was validated using NDVI values in absence of field data. NDVI and aquaculture fraction are positively correlated (R2 = 0.91), indicating the reliability of the sub-pixel classification