Abstract-In this paper, the use of Independent Component (IC) Discriminant Analysis (ICDA) for remote sensing classification is proposed. ICDA is a nonparametric method for discriminant analysis based on the application of a Bayesian classification rule on a signal composed by ICs. The method uses IC Analysis (ICA) to choose a transform matrix so that the transformed components are as independent as possible. When the data are projected in an independent space, the estimates of their multivariate density function can be computed in a much easier way as the product of univariate densities. A nonparametric kernel density estimator is used to compute the density functions of each IC. Finally, the Bayes rule is applied for the classification assignment. In this paper, we investigate the possibility of using ICDA for the classification of hyperspectral images. We study the influence of the algorithm used to enforce independence and of the number of IC retained for the classification, proposing an effective method to estimate the most suitable number. The proposed method is applied to several hyperspectral images, in order to test different data set conditions (urban/agricultural area, size of the training set, and type of sensor). Obtained results are compared with one of the most commonly used classifier of hyperspectral images (support vector machines) and show the comparative effectiveness of the proposed method in terms of accuracy.Index Terms-Bayesian classification, curse of dimensionality, hyperspectral data, Independent Component (IC) Analysis (ICA).
The identification of organic pollutants in the soil and the subsurface is a goal of primary importance in the management of contaminated sites. However, only a few non-invasive techniques can be useful towards this goal. One such technique is spectral induced polarization. In this study, we investigate the spectral induced polarization effect of changing fluid saturation in a well-characterized porous medium, analysing the difference between air and hydrocarbons, at different degrees of water saturation. The experiments were conducted on fine colic sand samples coming from an experimental site near Turin, Italy. Octanol and benzene were used as non-aqueous phase liquids. Samples were initially saturated with water having controlled electrical conductivity and Subsequently de-saturated stepwise with injection of air at known pressure. The colic sand samples were then re-saturated with the same water contaminated with hydrocarbons and then a non-aqueous phase liquid phase (either octanol or benzene) was injected in volumetric steps, in order to compare the effects of air and non-aqueous phase liquid invasion. At each saturation step, spectral induced polarization measurements were conducted in the 0.01 Hz to 1 kHz range using the ZEL-SIP04 impedance meter developed at the Forschungszentrum Juelich. The measurement setup guaranteed a 1 mrad phase precision for the entire frequency measurement range. Measurements were conducted under temperature controlled conditions at 20 (+/- 0.5)degrees C. All spectral induced polarization curves show a peak in the range 0.01-1 Hz that changes in intensity and frequency with varying saturation and a high-frequency phase shift increase dominated by capacitive coupling effects of the measuring system. A Multiple Cole-Cole model was fitted to the data. The effects of de-saturation on the low-frequency Cole-Cole parameters are that a) resistivity increases with decreasing water saturation but increases less with non-aqueous phase liquid than with the same volume of air; b) chargeability increases with decreasing water saturation but in presence of non-aqueous phase liquids its value is sometimes lower, sometimes higher and sometimes similar to the one observed in presence of air; c) the time constant tau increases with decreasing water saturation and is consistently larger with non-aqueous phase liquid than with air. These differences between air and non-aqueous phase liquid injection can be explained in terms of differences in non-aqueous phase distribution within the porous medium, as observed by X-ray micro-CT: while air is homogeneously distributed, non-aqueous phase liquids segregate under density effects. In summary, all spectral induced polarization effects of air and non-aqueous phase liquid injection in the considered porous medium are volumetric, i.e., are not due to interaction with grain surfaces or other electrical-chemical effects but are caused by pore obstruction by the electrically non-conductive phase
Catchment and hillslope hydrology is a major research area in geoscience and the understanding of its underlying processes is still poor. Direct investigation of steep hillslopes via drilling is often infeasible. In this paper, we present the results of non-invasive time-lapse monitoring of a controlled infiltration test at a site in the Italian Central Alps. The hillslope considered is steep (30-35 degrees), covered with grass and a soil layer 1-1.5 m thick above a variably fractured metamorphic bedrock. The key hydrologic question is whether rainfall infiltrates mainly into the underlying fractured bedrock, thus recharging a deeper hydraulic system, or flows in the soil layer as interflow towards the stream channel a few hundred metres downhill. In order to respond to this question, we applied 2200 mm of artificial rain on a 2 m x 2 m slope box over about 18 hours. We estimated the effective infiltration by subtracting the measured runoff (7% of total). Due to the limited irrigation time and the climate conditions, the evapotranspiration was considered as negligible. The soil moisture variation and the underlying bedrock were monitored via a combination of electrical resistivity tomography (ERT), TDR probes and tensiometers. A small-scale 3D cross-hole ERT experiment was performed using 2 m deep boreholes purposely drilled and completed with electrodes in the irrigated plot. A larger scale (35 m long) 2D surface ERT survey was also continuously acquired across the irrigated area. Monitoring continued up to 10 days after the experiment. As a result, we observed a very fast vertical infiltration through the soil cover, also favoured by preferential flow patterns, immediately followed by infiltration into the fractured bedrock. The surface layer showed a fast recovery of initial moisture condition nearly completed in the first 12 hours after the end of irrigation. The lateral transmission of infiltrating water and runoff were negligible as compared to the vertical infiltration. These experiment results confirm that the fractured bedrock has a key role in controlling the fast hydrological dynamics of the small catchment system under study. We concluded that deep water circulation is the key pathway to hillslope processes at this site
The Plator rock glacier is the first such landform identified in the Italian Alps that shows destabilised behaviour. Analysis of six sets of sequential orthophotographs from 1981 to 2012 reveals an exceptional advance of the rock glacier front (92.1 m) and a horizontal velocity up to 4 m a‐1 in different zones. The spatial variability of kinematics was evaluated by tracking sets of ‘tracer’ boulders on the rock glacier through time. Its velocity has progressively increased from the rooting zone to the tongue, with complex trends associated with distinct morphological features. Destabilisation likely occurred between 1954 and 1981, probably due to the relatively low elevation of the tongue, which resulted in warm permafrost conditions. Field observations reveal the presence of a large rock fall deposit, which occurred before 1981, and suggest that the debris overload could have triggered destabilisation. Since June 2015, an intensive monitoring programme has been implemented on the rock glacier, as the tongue is expected to travel over a steeper slope segment within the next 3 to 5 years, which could evolve in a catastrophic movement. Copyright © 2016 John Wiley & Sons, Ltd.
sification of hyperspectral images by using extended morphological attribute profiles and independent component analysis. Abstract-In this letter, a technique based on independent component analysis (ICA) and extended morphological attribute profiles (EAPs) is presented for the classification of hyperspectral images. The ICA maps the data into a subspace in which the components are as independent as possible. APs, which are extracted by using several attributes, are applied to each image associated with an extracted independent component, leading to a set of extended EAPs. Two approaches are presented for including the computed profiles in the analysis. The features extracted by the morphological processing are then classified with an SVM. The experiments carried out on two hyperspectral images proved the effectiveness of the proposed technique. Index Terms-Attribute filters, decision fusion, extended attribute profile (EAP), independent component analysis (ICA), mathematical morphology, remote sensing.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.