Vertex Component Analysis Research Articles

SPECTRAL UNMIXING ANALYSIS OF TIME SERIES LANDSAT 8 IMAGES

Abstract. Temporal analysis of Landsat 8 images opens up new opportunities in the unmixing procedure. Although spectral analysis of time series Landsat imagery has its own advantage, it has rarely been studied. Nevertheless, using the temporal information can provide improved unmixing performance when compared to independent image analyses. Moreover, different land cover types may demonstrate different temporal patterns, which can aid the discrimination of different natures. Therefore, this letter presents time series K-P-Means, a new solution to the problem of unmixing time series Landsat imagery. The proposed approach is to obtain the “purified” pixels in order to achieve optimal unmixing performance. The vertex component analysis (VCA) is used to extract endmembers for endmember initialization. First, nonnegative least square (NNLS) is used to estimate abundance maps by using the endmember. Then, the estimated endmember is the mean value of “purified” pixels, which is the residual of the mixed pixel after excluding the contribution of all nondominant endmembers. Assembling two main steps (abundance estimation and endmember update) into the iterative optimization framework generates the complete algorithm. Experiments using both simulated and real Landsat 8 images show that the proposed “joint unmixing” approach provides more accurate endmember and abundance estimation results compared with “separate unmixing” approach.

Comparison of Three Un-mixing Based Hyperspectral Image Fusion Techniques

ABSTRACT With the availability of multi sensor data in many fields, such as remote sensing, medical imaging or machine vision, sensor fusion has emerged as a new and promising research area. It is possible to have several images of the same scene providing different information although the scene is the same. This is because each image has been captured with a different sensor. A non-negative matrix factorization (NNMF) un mixing based fusion technique with vertex component analysis (VCA) based end member initialization and simple multiplicative update to improve the spatial resolution and to preserve the spectral resolution of the hyper spectral image is proposed. Its performance is analyzed with different number of iterations and end member initializations. A Constrained Non Negative Matrix Factorization unmixing based fusion technique is developed by adding a regularization term to the objective function to preserve the spectral resolution of the hyper spectral image, and its performance is analyzed with different number of iterations and end members. A rank two NNMF and hierarchical clustering based end member initialization and block principal pivoting algorithm based abundance estimation technique, for fusing hyper spectral image and simulated multispectral image is proposed and its performance is analyzed for different overlapping and non overlapping group of multispectral and hyper spectral bands. The performance of the above three methods are compared and analyzed. The obtained results show that the performance of rank two NNMF hierarchical clustering based fusion technique is better than the other two constrained and unconstrained NNMF un mixing based techniques. Also, the performance of these three proposed multi sensor image fusion techniques are compared with an existing image fusion technique.

Application of spatially related MRF model in NMF hyperspectral unmixing

Aiming at Non-negative Matrix Factorization (NMF)’s problem of initialization and "local minima" in hyperspectral unmixing, a NMF linear unmixing algorithm with spatial correlation constrains (SCNMF) based on Markov Random Field (MRF) was proposed. Firstly, Hyperspectral Signal identification by minimum error (HySime) method was adopted to estimate the number of endmembers, initialized endmember matrix and abundance matrix by Vertex Component Analysis (VCA) and Fully Constrained Least Squares (FCLS) respectively. then established energy function to depict the spatial distribution characteristics of ground objects by MRF model. Finally, spatial correlation constraint based on MRF model and NMF standard objective function were combined in the form of altemating iteration to estimate endmember spectrum and abundance of hyperspectral image. Theoretical analysis and experimental results indicated that, the endmember decomposition precision of SCNMF is 10.6% higher than that of Minimum Volume Constrained NMF (MVC-NMF), 12.3% higher than that of Piecewise Smoothness NMF with Sparseness Constraints(PSNMFSC), 14.1% higher than that of NMF with Alternating Projected Subgradients(APS-NMF); the abundance decomposition precision of SCNMF is 14.4% higher than that of MVC-NMF, 15.9% higher than that of PSNMFSC, 15.3% higher than that of APS-NMF.The proposed SCNMF can remedy NMF's deficiency in describing spatial correlation characteristics, and decrease spatial energy distribution error.

Pure endmember extraction using robust kernel archetypoid analysis for hyperspectral imagery

A robust kernel archetypoid analysis (RKADA) method is proposed to extract pure endmembers from hyperspectral imagery (HSI). The RKADA assumes that each pixel is a sparse linear mixture of all endmembers and each endmember corresponds to a real pixel in the image scene. First, it improves the re8gular archetypal analysis with a new binary sparse constraint, and the adoption of the kernel function constructs the principal convex hull in an infinite Hilbert space and enlarges the divergences between pairwise pixels. Second, the RKADA transfers the pure endmember extraction problem into an optimization problem by minimizing residual errors with the Huber loss function. The Huber loss function reduces the effects from big noises and outliers in the convergence procedure of RKADA and enhances the robustness of the optimization function. Third, the random kernel sinks for fast kernel matrix approximation and the two-stage algorithm for optimizing initial pure endmembers are utilized to improve its computational efficiency in realistic implementations of RKADA, respectively. The optimization equation of RKADA is solved by using the block coordinate descend scheme and the desired pure endmembers are finally obtained. Six state-of-the-art pure endmember extraction methods are employed to make comparisons with the RKADA on both synthetic and real Cuprite HSI datasets, including three geometrical algorithms vertex component analysis (VCA), alternative volume maximization (AVMAX) and orthogonal subspace projection (OSP), and three matrix factorization algorithms the preconditioning for successive projection algorithm (PreSPA), hierarchical clustering based on rank-two nonnegative matrix factorization (H2NMF) and self-dictionary multiple measurement vector (SDMMV). Experimental results show that the RKADA outperforms all the six methods in terms of spectral angle distance (SAD) and root-mean-square-error (RMSE). Moreover, the RKADA has short computational times in offline operations and shows significant improvement in identifying pure endmembers for ground objects with smaller spectrum differences. Therefore, the RKADA could be an alternative for pure endmember extraction from hyperspectral images.

Hybrid Spectral Unmixing: Using Artificial Neural Networks for Linear/Non-Linear Switching

Spectral unmixing is a key process in identifying spectral signature of materials and quantifying their spatial distribution over an image. The linear model is expected to provide acceptable results when two assumptions are satisfied: (1) The mixing process should occur at macroscopic level and (2) Photons must interact with single material before reaching the sensor. However, these assumptions do not always hold and more complex nonlinear models are required. This study proposes a new hybrid method for switching between linear and nonlinear spectral unmixing of hyperspectral data based on artificial neural networks. The neural networks was trained with parameters within a window of the pixel under consideration. These parameters are computed to represent the diversity of the neighboring pixels and are based on the Spectral Angular Distance, Covariance and a non linearity parameter. The endmembers were extracted using Vertex Component Analysis while the abundances were estimated using the method identified by the neural networks (Vertex Component Analysis, Fully Constraint Least Square Method, Polynomial Post Nonlinear Mixing Model or Generalized Bilinear Model). Results show that the hybrid method performs better than each of the individual techniques with high overall accuracy, while the abundance estimation error is significantly lower than that obtained using the individual methods. Experiments on both synthetic dataset and real hyperspectral images demonstrated that the proposed hybrid switch method is efficient for solving spectral unmixing of hyperspectral images as compared to individual algorithms.

Unmixing hyperspectral data by using signal subspace sampling

This paper demonstrates how Signal Subspace Sampling (SSS) is an effective pre-processing step for Non-negative Matrix Factorization (NMF) or Vertex Component Analysis (VCA). The approach allows to uniquely extract non-negative source signals which are orthogonal in at least one observation channel, respectively. It is thus well suited for processing hyperspectral images from X-ray microscopy, or other emission spectroscopies, into its non-negative source components. The key idea is to resample the given data so as to satisfy better the necessity and sufficiency conditions for the subsequent NMF or VCA. Results obtained both on an artificial simulation study as well as based on experimental data from electron-microscopy are reported.

The usage of data compression for the background estimation of electron energy loss spectra

Quantitative analysis of noisy electron spectrum images requires a robust estimation of the underlying background signal. We demonstrate how modern data compression methods can be used as a tool for achieving an analysis result less affected by statistical errors or to speed up the background estimation. In particular, we demonstrate how a multilinear singular value decomposition (MLSVD) can be used to enhance elemental maps obtained from a complex sample measured with energy electron loss spectroscopy. Furthermore, the usage of vertex component analysis (VCA) for a basis vector centered estimation of the background is demonstrated. Arising computational benefits in terms of model accuracy and computational costs are studied.

R-VCANet: A New Deep-Learning-Based Hyperspectral Image Classification Method

Deep-learning-based methods have displayed promising performance for hyperspectral image (HSI) classification, due to their capacity of extracting deep features from HSI. However, these methods usually require a large number of training samples. It is quite difficult for deep-learning model to provide representative feature expression for HSI data when the number of samples are limited. In this paper, a novel simplified deep-learning model, rolling guidance filter (RGF) and vertex component analysis network (R-VCANet), is proposed, which achieves higher accuracy when the number of training samples is not abundant. In R-VCANet, the inherent properties of HSI data, spatial information and spectral characteristics, are utilized to construct the network. And by this means the obtained model could generate more powerful feature expression with less samples. First, spectral and spatial information are combined via the RGF, which could explore the contextual structure features and remove small details from HSI. More importantly, we have designed a new network called vertex component analysis network for deep features extraction from the smoothed HSI. Experiments on three popular datasets indicate that the proposed R-VCANet based method reveals better performance than some state-of-the-art methods, especially when the training samples available are not abundant.

A machine learning framework to analyze hyperspectral stimulated Raman scattering microscopy images of expressed human meibum.

We develop and discuss a methodology for batch-level analysis of hyperspectral stimulated Raman scattering (hsSRS) data sets of human meibum in the CH-stretching vibrational range. The analysis consists of two steps. The first step uses a training set (n=19) to determine chemically meaningful reference spectra that jointly constitute a basis set for the sample. This procedure makes use of batch-level vertex component analysis (VCA), followed by unsupervised k-means clustering to express the data set in terms of spectra that represent lipid and protein mixtures in changing proportions. The second step uses a random forest classifier to rapidly classify hsSRS stacks in terms of the pre-determined basis set. The overall procedure allows a rapid quantitative analysis of large hsSRS data sets, enabling a direct comparison among samples using a single set of reference spectra. We apply this procedure to assess 50 specimens of expressed human meibum, rich in both protein and lipid, and show that the batch-level analysis reveals marked variation among samples that potentially correlate with meibum health quality.

End-member extraction based on segmented vertex component analysis in hyperspectral images.

Hyperspectral images collected by a remote sensing hyperspectral imaging instrument have many mixed pixels, due to the limited resolution of image sensors and the complex diversity of nature. End-member extraction is the process that determines the end-members in mixed pixels. The results of traditional methods are inaccurate, due to the spatial complexity and noise of actual hyperspectral image data. This study presents segmented vertex component analysis (SVCA), wherein the relative complexities of hyperspectral images are segmented into a number of relatively simple spatial subsets to reduce the effect of uncorrelated pixels. The end-members are extracted by finding the vertices of the simplex that minimally encloses the hyperspectral image data in each spatial subset, and the inversion abundance is used to identify each major end-member in each subset. Experimental results demonstrate that the proposed method can effectively implement end-member extraction with high accuracy.

Endmember extraction from hyperspectral image based on discrete firefly algorithm (EE-DFA)

This study proposed a novel method to extract endmembers from hyperspectral image based on discrete firefly algorithm (EE-DFA). Endmembers are the input of many spectral unmixing algorithms. Hence, in this paper, endmember extraction from hyperspectral image is regarded as a combinational optimization problem to get best spectral unmixing results, which can be solved by the discrete firefly algorithm. Two series of experiments were conducted on the synthetic hyperspectral datasets with different SNR and the AVIRIS Cuprite dataset, respectively. The experimental results were compared with the endmembers extracted by four popular methods: the sequential maximum angle convex cone (SMACC), N-FINDR, Vertex Component Analysis (VCA), and Minimum Volume Constrained Nonnegative Matrix Factorization (MVC-NMF). What’s more, the effect of the parameters in the proposed method was tested on both synthetic hyperspectral datasets and AVIRIS Cuprite dataset, and the recommended parameters setting was proposed. The results in this study demonstrated that the proposed EE-DFA method showed better performance than the existing popular methods. Moreover, EE-DFA is robust under different SNR conditions.

A Mutation Operator Accelerated Quantum-Behaved Particle Swarm Optimization Algorithm for Hyperspectral Endmember Extraction

The endmember extraction algorithm, which selects a collection of pure signature spectra for different materials, plays an important role in hyperspectral unmixing. In this paper, the endmember extraction algorithm is described as a combinatorial optimization problem and a novel Mutation Operator Accelerated Quantum-Behaved Particle Swarm Optimization (MOAQPSO) algorithm is proposed. The proposed approach employs Quantum-Behaved Particle Swarm Optimization (QPSO) to find endmembers with good performances. To the best of our knowledge, this is the first time that QPSO has been introduced into hyperspectral endmember extraction. In order to follow the law of particle movement, a high-dimensional particle definition is proposed. In addition, in order to avoid falling into a local optimum, a mutation operation is used to increase the population diversity. The proposed MOAQPSO algorithm was evaluated on both synthetic and real hyperspectral data sets. The experimental results indicated that the proposed method obtained better results than other state-of-the-art algorithms, including Vertex Component Analysis (VCA), N-FINDR, and Discrete Particle Swarm Optimization (D-PSO).

Analysis of electron energy loss spectroscopy data using geometric extraction methods

A set of geometric data decomposition methods is discussed. In particular, randomized vertex component analysis (RVCA), an extension of vertex component analysis (VCA) for the application to noisy data, is established. Minimum volume simplex analysis (MVSA), a recent technique for the extraction of endmembers in the absence of pure pixels, is presented. A comparison between MVSA and the previously presented technique of Bayesian Linear Unmixing (BLU) is drawn. Lastly, the efficiency of these methods for high-dimensional data is examined. Improvement on the extracted source components spectral signatures are achieved by establishing Gaussian mixture modeling as extraction technique.

Comparative Study and Analysis Among ATGP, VCA, and SGA for Finding Endmembers in Hyperspectral Imagery

Endmember finding has become increasingly important in hyperspectral data exploitation because endmembers can be used to specify unknown particular spectral classes. Pixel purity index (PPI) and N-finder algorithm (N-FINDR) are probably the two most widely used techniques for this purpose where many currently available endmember finding algorithms are indeed derived from these two algorithms and can be considered as their variants. Among them are three well-known algorithms derived from imposing different abundance constraints, that is, abundance-unconstrained automatic target generation process (ATGP), abundance nonnegativity constrained vertex component analysis (VCA), and fully abundance constrained simplex growing algorithm (SGA). This paper explores relationships among these three algorithms and further shows that theoretically they are essentially the same algorithms in the sense of design rationale. The reason that these three algorithms perform differently is not because they are different algorithms, but rather because they use different preprocessing steps, such as initial conditions and dimensionality reduction transforms.

Locality Preserving Projection Based on Endmember Extraction for Hyperspectral Image Dimensionality Reduction and Target Detection.

In order to reduce the effect of spectral variability on calculation precision for the weighted matrix in the locality preserving projection (LPP) algorithm, an improved dimensionality reduction method named endmember extraction-based locality preserving projection (EE-LPP) is proposed in this paper. The method primarily uses the vertex component analysis (VCA) method to extract endmember spectra from hyperspectral imagery. It then calculates the similarity between pixel spectra and the endmember spectra by using the spectral angle distance, and uses it as the basis for selecting neighboring pixels in the image and constructs a weighted matrix between pixels. Finally, based on the weighted matrix, the idea of the LPP algorithm is applied to reduce the dimensions of hyperspectral image data. Experimental results of real hyperspectral data demonstrate that the low-dimensional features acquired by the proposed methods can fully reflect the characteristics of the original image and further improve target detection accuracy.

Endmember initialization method for hyperspectral data unmixing

Spectral unmixing aims at finding the spectrally pure constituent materials (also called endmembers) and their respective fractional abundances in each pixel of the hyperspectral image scene. One important issue in hyperspectral data unmixing is the initialization of endmembers. Most unmixing methods initialize their endmembers by randomly selecting a specified number of pixels from the data or by vertex component analysis, which limits their performance in practice. We propose an endmember initialization method for hyperspectral data unmixing. Our initial endmembers include some of the true endmembers, which improves the accuracy of hyperpspectral unmixing effectively. The experimental results on both synthetic and real hyperspectral data illustrate the superiority of the proposed method compared with other state-of-the-art approaches.

Abstract 4246: Multispectral open-air fluorescence-guided imaging and detection of tumors using a hands-free translational platform with liquid crystal tunable filters (LCTF)

Abstract Intraoperative identification and resection of tumors currently relies on the ability of the surgeon to visually detect or palpate the tumors and residual malignant tissue. As such, minuscule tumor nodules can go undetected or be inadequately removed, with such cases often resulting in the need for secondary treatment or additional surgical intervention. The Solaris™ platform is an open-air fluorescent imaging instrument designed for large animal fluorescence-guided surgery, with the advantage of real-time acquisition of fluorescence images/video under surgical light conditions. Solaris supports four fixed fluorescent channels ranging from visible to near infrared (NIR), and a multispectral channel where a liquid crystal tunable filter (LCTF) is used to acquire spectral data by sweeping across the green-to-red portion of the visible spectrum. This range of imaging channels allows for single-wavelength and multispectral imaging of widely used reagents (e.g. indocyanine green [ICG] and Fluorescein isothiocyanate [FITC]) and unique NIR fluorescent dyes used for detecting and labeling tumors. While fluorescent imaging using NIR imaging agents (680, 750, 800 nm) offered effective tumor detection, identification of tumors implanted in nude mice or rats using visible (400-650 nm) reagents such as FITC presented challenges considering the presence of auto-fluorescence originating from tissue and food (alfalfa). For these reagents, Solaris acquired multispectral images using the LCTF under ambient light conditions, and an automated spectral unmixing algorithm was applied to the multispectral data, after background correction and ambient light removal, to separate tissue and food auto-fluorescence from the reagent fluorescent signal. The algorithm used vertex component analysis to automatically extract the primary pure spectra present in the multispectral images and unmix the reagent fluorescent signal by non-negative least squares fitting. To test the spectral unmixing capabilities of Solaris, in vivo experiments were performed using small amounts of locally injected FITC in mice and rats. In the absence of unmixing, it was not possible to accurately detect sites of FITC signal, but with unmixing the labeled regions were well defined. Additional studies in tumor-bearing mice and rats substantiated the ability to spectrally unmix FITC agent signal in deep tumor masses imaged under ambient light, enhancing the ability to surgically resect them. To further validate this concept, bioluminescent tumor cell lines were implanted in mice. After image-guided tumor resection, both the residual tumor bed and the resected tumors were imaged to confirm complete removal. These data demonstrate that intraoperative image-guided resection of fluorescent-labeled tumors can be achieved using LCTF-based open-air multispectral imaging on the Solaris. Citation Format: Ali Behrooz, Kristine Vasquez, Peter Waterman, Jeff Meganck, Jeffrey Peterson, Peter Miller, Joshua Kempner, Wael Yared. Multispectral open-air fluorescence-guided imaging and detection of tumors using a hands-free translational platform with liquid crystal tunable filters (LCTF). [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 4246.

Clustering based band selection for endmember extraction using simplex growing algorithm in hyperspectral images

With the advancement in technology, hyperspectral images have potential applications in the field of remote sensing due to their high spectral resolution. Despite the hyperspectral image providing abundant information, its analysis suffers from the problem of high dimensionality. Hence, Dimensionality Reduction (DR) is an essential task in all hyperspectral image analysis. Band Selection, which is one of the DR techniques, is still a challenging issue even though many algorithms have been developed. To provide remedy for this issue, this paper explores a novel approach for band selection using K-means clustering on statistical feature in hyperspectral images. The proposed method of clustering based band selection for DR is simple and accurate. A reliable estimate of number of bands to be selected is provided by Virtual Dimensionality (VD). Informative bands preserving maximum information are selected based on the statistical feature, the variance using K-means Clustering technique. Further, our proposed work involves the utilization of the effectiveness of Simplex Growing Algorithm (SGA) on endmember extraction in association with clustering based band selection. Using Fully Constrained Least Squares (FCLS) method, abundance fraction is estimated based on endmember signatures, which are derived using Endmember Extraction Algorithm (EEA). The proposed work is investigated and compared with that of N-FINDR and Vertex Component Analysis (VCA) algorithms. The performance of the proposed algorithm is evaluated using Root Mean Square Error (RMSE), Spectral Angle Distance (SAD) and computation time. Experimental results show that the proposed clustering based band selection with SGA endmember extraction algorithm reduces the average SAD by 8 to 10 % and the average RMSE by nearly 1 %, compared to that of N-FINDR and VCA algorithms. In terms of computation time, the proposed band selection based DR with SGA algorithm is seven times faster than conventional transform based DR with SGA algorithm.

Waterproofing in Arabidopsis: Following Phenolics and Lipids In situ by Confocal Raman Microscopy.

Waterproofing of the aerial organs of plants imposed a big evolutionary step during the colonization of the terrestrial environment. The main plant polymers responsible of water repelling are lipids and lignin, which play also important roles in the protection against biotic/abiotic stresses, regulation of flux of gases and solutes, and mechanical stability against negative pressure, among others. While the lipids, non-polymerized cuticular waxes together with the polymerized cutin, protect the outer surface, lignin is confined to the secondary cell wall within mechanical important tissues. In the present work a micro cross-section of the stem of Arabidopsis thaliana was used to track in situ the distribution of these non-carbohydrate polymers by Confocal Raman Microscopy. Raman hyperspectral imaging gives a molecular fingerprint of the native waterproofing tissues and cells with diffraction limited spatial resolution (~300 nm) at relatively high speed and without any tedious sample preparation. Lipids and lignified tissues as well as their effect on water content was directly visualized by integrating the 1299, 1600, and 3400 cm−1 band, respectively. For detailed insights into compositional changes of these polymers vertex component analysis was performed on selected sample positions. Changes have been elucidated in the composition of lignin within the lignified tissues and between interfascicular fibers and xylem vessels. Hydrophobizing changes were revealed from the epidermal layer to the cuticle as well as a change in the aromatic composition within the cuticle of trichomes. To verify Raman signatures of different waterproofing polymers additionally Raman spectra of the cuticle and cutin monomer from tomato (Solanum lycopersicum) as well as aromatic model polymers (milled wood lignin and dehydrogenation polymer of coniferyl alcohol) and phenolic acids were acquired.

A Hybrid CPU–GPU Real-Time Hyperspectral Unmixing Chain

Hyperspectral images are used in different applications in Earth and space science, and many of these applications exhibit real- or near real-time constraints. A problem when analyzing hyperspectral images is that their spatial resolution is generally not enough to separate different spectrally pure constituents (endmembers); as a result, several of them can be found in the same pixel. Spectral unmixing is an important technique for hyperspectral data exploitation, aimed at finding the spectral signatures of the endmembers and their associated abundance fractions. The development of techniques able to provide unmixing results in real-time is a long desired goal in the hyperspectral imaging community. In this paper, we describe a real-time hyperspectral unmixing chain based on three main steps: 1) estimation of the number of endmembers using the hyperspectral subspace identification with minimum error (HySime); 2) estimation of the spectral signatures of the endmembers using the vertex component analysis (VCA); and 3) unconstrained abundance estimation. We have developed new parallel implementations of the aforementioned algorithms and assembled them in a fully operative real-time unmixing chain using graphics processing units (GPUs), exploiting NVIDIA’s compute unified device architecture (CUDA) and its basic linear algebra subroutines (CuBLAS) library, as well as OpenMP and BLAS for multicore parallelization. As a result, our real-time chain exploits both CPU (multicore) and GPU paradigms in the optimization. Our experiments reveal that this hybrid GPU–CPU parallel implementation fully meets real-time constraints in hyperspectral imaging applications.