Vertex component analysis: a geometric-based approach to unmix hyperspectral data

Hyperspectral remote sensing exploits the electromagnetic scattering patterns of the different materials at specific wavelengths [2, 3]. Hyperspectral sensors have been developed to sample the scattered portion of the electromagnetic spectrum extending from the visible region through the near-infrared and mid-infrared, in hundreds of narrow contiguous bands [4, 5]. The number and variety of potential civilian and military applications of hyperspectral remote sensing is enormous [6, 7]. Very often, the resolution cell corresponding to a single pixel in an image contains several substances (endmembers) [4]. In this situation, the scattered energy is a mixing of the endmember spectra. A challenging task underlying many hyperspectral imagery applications is then decomposing a mixed pixel into a collection of reflectance spectra, called endmember signatures, and the corresponding abundance fractions [8–10]. Depending on the mixing scales at each pixel, the observed mixture is either linear or nonlinear [11, 12]. Linear mixing model holds approximately when the mixing scale is macroscopic [13] and there is negligible interaction among distinct endmembers [3, 14]. If, however, the mixing scale is microscopic (or intimate mixtures) [15, 16] and the incident solar radiation is scattered by the scene through multiple bounces involving several endmembers [17], the linear model is no longer accurate. Linear spectral unmixing has been intensively researched in the last years [9, 10, 12, 18–21]. It considers that a mixed pixel is a linear combination of endmember signatures weighted by the correspondent abundance fractions. Under this model, and assuming that the number of substances and their reflectance spectra are known, hyperspectral unmixing is a linear problem for which many solutions have been proposed (e.g., maximum likelihood estimation [8], spectral signature matching [22], spectral angle mapper [23], subspace projection methods [24,25], and constrained least squares [26]). In most cases, the number of substances and their reflectances are not known and, then, hyperspectral unmixing falls into the class of blind source separation problems [27]. Independent component analysis (ICA) has recently been proposed as a tool to blindly unmix hyperspectral data [28–31]. ICA is based on the assumption of mutually independent sources (abundance fractions), which is not the case of hyperspectral data, since the sum of abundance fractions is constant, implying statistical dependence among them. This dependence compromises ICA applicability to hyperspectral images as shown in Refs. [21, 32]. In fact, ICA finds the endmember signatures by multiplying the spectral vectors with an unmixing matrix, which minimizes the mutual information among sources. If sources are independent, ICA provides the correct unmixing, since the minimum of the mutual information is obtained only when sources are independent. This is no longer true for dependent abundance fractions. Nevertheless, some endmembers may be approximately unmixed. These aspects are addressed in Ref. [33]. Under the linear mixing model, the observations from a scene are in a simplex whose vertices correspond to the endmembers. Several approaches [34–36] have exploited this geometric feature of hyperspectral mixtures [35]. Minimum volume transform (MVT) algorithm [36] determines the simplex of minimum volume containing the data. The method presented in Ref. [37] is also of MVT type but, by introducing the notion of bundles, it takes into account the endmember variability usually present in hyperspectral mixtures. The MVT type approaches are complex from the computational point of view. Usually, these algorithms find in the first place the convex hull defined by the observed data and then fit a minimum volume simplex to it. For example, the gift wrapping algorithm [38] computes the convex hull of n data points in a d-dimensional space with a computational complexity of O(nbd=2cþ1), where bxc is the highest integer lower or equal than x and n is the number of samples. The complexity of the method presented in Ref. [37] is even higher, since the temperature of the simulated annealing algorithm used shall follow a log( ) law [39] to assure convergence (in probability) to the desired solution. Aiming at a lower computational complexity, some algorithms such as the pixel purity index (PPI) [35] and the N-FINDR [40] still find the minimum volume simplex containing the data cloud, but they assume the presence of at least one pure pixel of each endmember in the data. This is a strong requisite that may not hold in some data sets. In any case, these algorithms find the set of most pure pixels in the data. PPI algorithm uses the minimum noise fraction (MNF) [41] as a preprocessing step to reduce dimensionality and to improve the signal-to-noise ratio (SNR). The algorithm then projects every spectral vector onto skewers (large number of random vectors) [35, 42,43]. The points corresponding to extremes, for each skewer direction, are stored. A cumulative account records the number of times each pixel (i.e., a given spectral vector) is found to be an extreme. The pixels with the highest scores are the purest ones. N-FINDR algorithm [40] is based on the fact that in p spectral dimensions, the p-volume defined by a simplex formed by the purest pixels is larger than any other volume defined by any other combination of pixels. This algorithm finds the set of pixels defining the largest volume by inflating a simplex inside the data. ORA SIS [44, 45] is a hyperspectral framework developed by the U.S. Naval Research Laboratory consisting of several algorithms organized in six modules: exemplar selector, adaptative learner, demixer, knowledge base or spectral library, and spatial postrocessor. The first step consists in flat-fielding the spectra. Next, the exemplar selection module is used to select spectral vectors that best represent the smaller convex cone containing the data. The other pixels are rejected when the spectral angle distance (SAD) is less than a given thresh old. The procedure finds the basis for a subspace of a lower dimension using a modified Gram–Schmidt orthogonalizati on. The selected vectors are then projected onto this subspace and a simplex is found by an MV T pro cess. ORA SIS is oriented to real-time target detection from uncrewed air vehicles using hyperspectral data [46]. In this chapter we develop a new algorithm to unmix linear mixtures of endmember spectra. First, the algorithm determines the number of endmembers and the signal subspace using a newly developed concept [47, 48]. Second, the algorithm extracts the most pure pixels present in the data. Unlike other methods, this algorithm is completely automatic and unsupervised. To estimate the number of endmembers and the signal subspace in hyperspectral linear mixtures, the proposed scheme begins by estimating sign al and noise correlation matrices. The latter is based on multiple regression theory. The signal subspace is then identified by selectin g the set of signal eigenvalue s that best represents the data, in the least-square sense [48,49 ], we note, however, that VCA works with projected and with unprojected data. The extraction of the end members exploits two facts: (1) the endmembers are the vertices of a simplex and (2) the affine transformation of a simplex is also a simplex. As PPI and N-FIND R algorithms, VCA also assumes the presence of pure pixels in the data. The algorithm iteratively projects data on to a direction orthogonal to the subspace spanned by the endmembers already determined. The new end member signature corresponds to the extreme of the projection. The algorithm iterates until all end members are exhausted. VCA performs much better than PPI and better than or comparable to N-FI NDR; yet it has a computational complexity between on e and two orders of magnitude lower than N-FINDR. The chapter is structure d as follows. Section 19.2 describes the fundamentals of the proposed method. Section 19.3 and Section 19.4 evaluate the proposed algorithm using simulated and real data, respectively. Section 19.5 presents some concluding remarks.

TIGRE - An autonomous ground robot for outdoor exploration

13th International Conference on Autonomous Robot Systems (Robotica), 2013

Calibration Method for Underwater Visual Ground-Truth System

This work presents an automatic calibration method for a vision based external underwater ground-truth positioning system. These systems are a relevant tool in benchmarking and assessing the quality of research in underwater robotics applications. A stereo vision system can in suitable environments such as test tanks or in clear water conditions provide accurate position with low cost and flexible operation. In this work we present a two step extrinsic camera parameter calibration procedure in order to reduce the setup time and provide accurate results. The proposed method uses a planar homography decomposition in order to determine the relative camera poses and the determination of vanishing points of detected lines in the image to obtain the global pose of the stereo rig in the reference frame. This method was applied to our external vision based ground-truth at the INESC TEC/Robotics test tank. Results are presented in comparison with an precise calibration performed using points obtained from an accurate 3D LIDAR modelling of the environment.

Uncertainty based Multi-Robot Cooperative Triangulation

The paper presents a multi-robot cooperative framework to estimate the 3D position of dynamic targets, based on bearing-only vision measurements. The uncertainty of the observation provided by each robot equipped with a bearing-only vision system is effectively addressed for cooperative triangulation purposes by weighing the contribution of each monocular bearing ray in a probabilistic manner. The envisioned framework is evaluated in an outdoor scenario with a team of heterogeneous robots composed of an Unmanned Ground and Aerial Vehicle.

Is land surface temperature an earthquake precursor?

Dissertation submitted in partial fulfillment of the requirements for the Degree of Master of Science in Geospatial Technologies

Recent droughts and the impact of North Atlantic Oscillation in Iberia; a spatiotemporal analysis based on vegetation temperature condition Index (VTCI)

Dissertation submitted in partial fulfillment of the requirements for the Degree of Master of Science in Geospatial Technologies.

Business planning in nonprofit: laying the ground for customer relationship management in Instituto de Empreendedorismo Social

A Work Project, presented as part of the requirements for the Award of a Masters Degree in Management from the NOVA – School of Business and Economics

Historical reconstructions of raw materials based on a blue smalt coating applied to a seventeenth-century altarpiece

Based on samples cross-sections from the Main Altarpiece of the Coimbra Old Cathedral, where a blue coating performed in 1685 is observed (that was partly covered with a Prussian blue-containing overpaint), the raw materials present in this coating were reproduced and studied. Blue areas were painted with smalt in oil, according to the contract signed by Manoel da Costa Pereira in 1684 and the analysis by Le Gac in 2009. Based on these, three batches of cobalt-based glasses (S1, S2 and S3) were heated and melted in alumina crucibles in the kiln. S1 contained 6.03 % of cobalt oxide, S2 contained 2.10 %, with the addition of 1.49 % of magnesium oxide, and S3 contained 6.82 % of cobalt oxide, with the addition of 4.63% of antimony trioxide. These batches were ground mechanically with water and manually with different vehicles stated in recipes. The results were studied by means of OM, SEM-EDS, X-Ray CT, Colorimetry and Vickers HT. Different binders were also produced and analyzed by means of μ-FTIR, in order to perform their characterization and obtain Standard Spectra. Since anhydrite was identified in the ground layers, gypsum from Óbidos was also characterized by XRD. The main goal of this thesis was to study all the raw materials present in the 1685-blue coating, in order to allow the historically accurate reconstruction of the layers build-up in the next future.

New self-assembled supramolecular hydrogels based on dehydropeptides

Supramolecular hydrogels rely on small molecules that self-assemble in water as a result of the cooperative effect of several relatively weak intermolecular interactions. Peptide-based low molecular weight hydrogelators have attracted enormous interest owing to the simplicity of small molecules combined with the versatility and biocompatibility of peptides. In this work, naproxen, a well known non-steroidal anti-inflammatory drug, was N-conjugated with various dehydrodipeptides to give aromatic peptide amphiphiles that resist proteolysis. Molecular dynamics simulations were used to obtain insight into the underlying molecular mechanism of self-assembly and to rationalize the design of this type of hydrogelators. The results obtained were in excellent agreement with the experimental observations. Only dehydrodipeptides having at least one aromatic amino acid gave hydrogels. The characterization of the hydrogels was carried out using transmission electron microscopy (TEM), circular dichroism (CD), fluorescence spectroscopy and also rheological assays.

Development of a data acquisition system to characterise sustainable energy technologies and to support Web-Based learning (DAQ-WBL)

Driven by concerns about rising energy costs, security of supply and climate change a new wave of Sustainable Energy Technologies (SET’s) have been embraced by the Irish consumer. Such systems as solar collectors, heat pumps and biomass boilers have become common due to government backed financial incentives and revisions of the building regulations. However, there is a deficit of knowledge and understanding of how these technologies operate and perform under Ireland’s maritime climate. This AQ-WBL project was designed to address both these needs by developing a Data Acquisition (DAQ) system to monitor the performance of such technologies and a web-based learning environment to disseminate performance characteristics and supplementary information about these systems. A DAQ system consisting of 108 sensors was developed as part of Galway-Mayo Institute of Technology’s (GMIT’s) Centre for the Integration of Sustainable EnergyTechnologies (CiSET) in an effort to benchmark the performance of solar thermal collectors and Ground Source Heat Pumps (GSHP’s) under Irish maritime climate, research new methods of integrating these systems within the built environment and raise awareness of SET’s. It has operated reliably for over 2 years and has acquired over 25 million data points. Raising awareness of these SET’s is carried out through the dissemination of the performance data through an online learning environment. A learning environment was created to provide different user groups with a basic understanding of a SET’s with the support of performance data, through a novel 5 step learning process and two examples were developed for the solar thermal collectors and the weather station which can be viewed at http://www.kdp 1 .aquaculture.ie/index.aspx. This online learning environment has been demonstrated to and well received by different groups of GMIT’s undergraduate students and plans have been made to develop it further to support education, awareness, research and regional development.

Observations on the morphology of Australorbis nigricans

A morphological study was done on A. nigricans, based on the observation of shell, radula, renal region and genitalia of 50 specimens measuring 18 mm in diameter. The data obtained are to be compared with those recorded in our previous paper (PARAENSE & DESLANDES, 1955) on A. glabratus. The characteristics common to both species will not be mentioned here. The numerals refere to the means and their standard deviations: no special reference being done, they correspond to length measurementes. Shell - 18 mm in diameter, 6.37 ± 0.29 mm in greatest width, 6 whorls. Prevailing colur ferruginous sepia, a minority of olivaceous, ochreous, nigrescent and deeply black specimens being found. Right side variously depressed, umbilicated, 1.5 to 3.5 mm deep from the bottom of the umblicus to the highest level of the last whorl. Left side more depressed than the right one, broadly concave, 1.5 to 3.5 mm deep. Both sides show a varously distinct keel, that looks sharper at the left. Aperture deltoid, varying in outline and width. Body, extended - 60.26 ± 3.62 mm, less pigmented than in glabratus. Renal tube - 30.68 ± 1.69 mm, showing neither ridge nor pigmented line along its ventral surface, this negative character affording a sure means of separation from glabratus. Ovotestis - 14.48 ± 1.93 mm. Ovisperm duct - 13.04 ± 1.60 mm, including the non-unwound seminal vesicle. The latter was 0.97 ± 0,21 mm in greatest width. Carrefour - Resembling that of glabratus. Sperm duct - 21.36 ± 1.53 mm. Prostate - Prostate duct 7.14 ± 0.74 mm, collecting a row of long diverticula numbering 19.6 ± 3.1 and more separate than in glabratus. Last diverticulum generally bifurcate or arborescent, the remaining ones arborescent. Vas deferens - 28.68 ± 1.38. Ratio vas deferens/vergic sac = 6.8±0.8. Verge - 3.08 ± 0.28 mm long, 0.11 ± 0.02 mm wide. Vergic sac - 3.07 ± 0.28 mm long, about 0.20 mm wide. Ratio vergic sac/preputium = 0.84 ± 0.12. Preputium - 3.69 ± 0.47 mm long, 0.85 ± 0.10 mm wide. Albumen gland - Resembling taht of glabratus. Oviduct - 16.26 ± 1.41 mm, swollen at the cephalic end. Uterus - 13.24 ± 1.19 mm. Vagina - 1.70 ± 0.22 mm, swolen at the caudal portion. Spermatheca - 2.78 ± 0.40 mm long, 0.86 ± 0.16 mm wide. Spermathecal duct 1.11 ± 0.20 mm. Radula - 125 to 168 horizontal rows of teeth (mean 153.9 ± 8.4). Radula formula 28-1-28 to 36-1-36 (mean 31.8 ± 1.9). Mode formula 31-1-31. The morphological characteristics of the renal region and shell, and the great body length in the same condition of shell diameter, distinguish A. nigricans from the most related species A. glabratus, giving support to considering it a good species from a txonomic or phenotypic standpoint (morphospecies).

Elevated blood pressure prevalence in children did not follow the increase in overweight: results from a school-based study in a rapidly developing country, 1998-2006

Background: Blood pressure (BP) is strongly associated with body weight and there is concern that the pediatric overweight epidemic could lead to an increase in children's mean BP. Objectives: We analyzed BP trends from 1998 to 2006 among children of the Seychelles, a rapidly developing middle-income country in Africa. Methods: Serial school-based surveys of weight, height and BP were conducted yearly between 1998-2006 among all students of the country in four school grades (kindergarten, 4th, 7th and 10th years of compulsory school). We used the CDC criteria to define "overweight" (BMI _95th sex-, and age-specific percentile) and the NHBPEP criteria for "elevated BP" (BP _95th sex-, age-, and height specific percentile). Methods for height, weight, and BP measurements were identical over the study period. The trends in mean BMI and mean systolic/diastolic BP were assessed with linear regression. Results: 27,703 children aged 4-18 years (participation rate: 79%) contributed 43,927 observations on weight, height, and BP. The prevalence of overweight increased from 5.1% in 1998-2000 to 8.1% in 2004-2006 among boys, and from 6.1% to 9.1% among girls, respectively. The prevalence of elevated BP was 8.4% in 1998-2000 and 6.9% in 2004-2006 among boys; 9.8% and 7.8% among girls, respectively. Over the 9-years study period, age-adjusted body mass index (BMI) increased by 0.078 kg/m2/year in boys and by 0.083 kg/m2/year in girls (both sexes, P_0.001). Age- and height-adjusted systolic BP decreased by -0.37 mmHg/year in boys and by -0.34 mmHg/year in girls (both sexes, P_0.001). Diastolic BP did not change in boys (-0.02 mmHg/year, P: 0.40) and slightly increased in girls (0.07 mmHg/year, P: 0.003). These trend estimates were altered modestly upon further adjustment for BMI or if analyses were based on median rather than mean values. Conclusion: Although body weight increased markedly between 1998 and 2006 in this population, systolic BP decreased and diastolic BP changed only marginally. This suggests that population increases in body weight are not necessarily associated with corresponding rises in BP in children.

A Modulated Closed Form solution for Quantitative Susceptibility Mapping - A thorough evaluation and comparison to iterative methods based on edge prior knowledge.

The aim of this study is to perform a thorough comparison of quantitative susceptibility mapping (QSM) techniques and their dependence on the assumptions made. The compared methodologies were: two iterative single orientation methodologies minimizing the l2, l1TV norm of the prior knowledge of the edges of the object, one over-determined multiple orientation method (COSMOS) and anewly proposed modulated closed-form solution (MCF). The performance of these methods was compared using a numerical phantom and in-vivo high resolution (0.65mm isotropic) brain data acquired at 7T using a new coil combination method. For all QSM methods, the relevant regularization and prior-knowledge parameters were systematically changed in order to evaluate the optimal reconstruction in the presence and absence of a ground truth. Additionally, the QSM contrast was compared to conventional gradient recalled echo (GRE) magnitude and R2* maps obtained from the same dataset. The QSM reconstruction results of the single orientation methods show comparable performance. The MCF method has the highest correlation (corrMCF=0.95, r(2)MCF =0.97) with the state of the art method (COSMOS) with additional advantage of extreme fast computation time. The l-curve method gave the visually most satisfactory balance between reduction of streaking artifacts and over-regularization with the latter being overemphasized when the using the COSMOS susceptibility maps as ground-truth. R2* and susceptibility maps, when calculated from the same datasets, although based on distinct features of the data, have a comparable ability to distinguish deep gray matter structures.

Turning sunlight into stone: the oxalate-carbonate pathway in a tropical tree ecosystem.

An African oxalogenic tree, the iroko tree (Milicia excelsa), has the property to enhance carbonate precipitation in tropical oxisols, where such accumulations are not expected due to the acidic conditions in these types of soils. This uncommon process is linked to the oxalate-carbonate pathway, which increases soil pH through oxalate oxidation. In order to investigate the oxalate-carbonate pathway in the iroko system, fluxes of matter have been identified, described, and evaluated from field to microscopic scales. In the first centimeters of the soil profile, decaying of the organic matter allows the release of whewellite crystals, mainly due to the action of termites and saprophytic fungi. In addition, a concomitant flux of carbonate formed in wood tissues contributes to the carbonate flux and is identified as a direct consequence of wood feeding by termites. Nevertheless, calcite biomineralization of the tree is not a consequence of in situ oxalate consumption, but rather related to the oxalate oxidation inside the upper part of the soil. The consequence of this oxidation is the presence of carbonate ions in the soil solution pumped through the roots, leading to preferential mineralization of the roots and the trunk base. An ideal scenario for the iroko biomineralization and soil carbonate accumulation starts with oxalatization: as the iroko tree grows, the organic matter flux to the soil constitutes the litter, and an oxalate pool is formed on the forest ground. Then, wood rotting agents (mainly termites, saprophytic fungi, and bacteria) release significant amounts of oxalate crystals from decaying plant tissues. In addition, some of these agents are themselves producers of oxalate (e.g. fungi). Both processes contribute to a soil pool of "available" oxalate crystals. Oxalate consumption by oxalotrophic bacteria can then start. Carbonate and calcium ions present in the soil solution represent the end products of the oxalate-carbonate pathway. The solution is pumped through the roots, leading to carbonate precipitation. The main pools of carbon are clearly identified as the organic matter (the tree and its organic products), the oxalate crystals, and the various carbonate features. A functional model based on field observations and diagenetic investigations with δ13C signatures of the various compartments involved in the local carbon cycle is proposed. It suggests that the iroko ecosystem can act as a long-term carbon sink, as long as the calcium source is related to non-carbonate rocks. Consequently, this carbon sink, driven by the oxalate carbonate pathway around an iroko tree, constitutes a true carbon trapping ecosystem as defined by ecological theory.

Semi-supervised segmentation of ultrasound images based on patch representation and continuous min cut.

Ultrasound segmentation is a challenging problem due to the inherent speckle and some artifacts like shadows, attenuation and signal dropout. Existing methods need to include strong priors like shape priors or analytical intensity models to succeed in the segmentation. However, such priors tend to limit these methods to a specific target or imaging settings, and they are not always applicable to pathological cases. This work introduces a semi-supervised segmentation framework for ultrasound imaging that alleviates the limitation of fully automatic segmentation, that is, it is applicable to any kind of target and imaging settings. Our methodology uses a graph of image patches to represent the ultrasound image and user-assisted initialization with labels, which acts as soft priors. The segmentation problem is formulated as a continuous minimum cut problem and solved with an efficient optimization algorithm. We validate our segmentation framework on clinical ultrasound imaging (prostate, fetus, and tumors of the liver and eye). We obtain high similarity agreement with the ground truth provided by medical expert delineations in all applications (94% DICE values in average) and the proposed algorithm performs favorably with the literature.