Edge-preserving image denoising and estimation of discontinuous surfaces

In this paper, we are interested in the problem of estimating a discontinuous surface from noisy data. A novel procedure for this problem is proposed based on local linear kernel smoothing, in which local neighborhoods are adapted to the local smoothness of the surface measured by the observed data. The procedure can therefore remove noise correctly in continuity regions of the surface and preserve discontinuities at the same time. Since an image can be regarded as a surface of the image intensity function and such a surface has discontinuities at the outlines of objects, this procedure can be applied directly to image denoising. Numerical studies show that it works well in applications, compared to some existing procedures.     

Contribution bacterienne a la corrosion et a la protection des metaux

The metabolism of aerobic bacteria and anaerobic bacteria being very different, their influence on the metal-corrosion is also not similar.A living bacterium, in nutrient media, at optimal pH and redox potential, grows very well and the corrosion disappears. The non-proteolytic, reductor and oxygen-using bacteria are the best. But, if a bacterium gives complexing products, it dies and the corrosion increases.Experiments in media satured with O₂ or N₂ have proved that a bacterium decreasing the corrosion in aerobic, increases the corrosion when it is put in anaerobic media.     

Conditions D'immunite Et De Corrosion De L'aluminium Dans Un Milieu Bacterien Doux Et Marin

No corrosion was observed (with or without protective film) when the metal was exposed to bacteria growing in full nutrient medium. Pitting corrosion occurred when the bacteria died or when their metabolism was disturbed by decomposition of the nutrient medium or by the presence of A1³⁺ ions.     

Kinetic studies of elemental mercury adsorption in activated carbon fixed bed reactor

Activated carbons are suitable materials for Hg(0) adsorption in fixed bed operation or in injection process. The fixed bed tests provide good indication of activated carbons effectiveness and service lives, which depend on the rates of Hg(0) adsorption. In order to correlate fixed bed properties and operation conditions, with their adsorptive capacity and saturation time, Hg(0) adsorption tests were realized in a bench-scale unit, consisted of F400 activated carbon fixed bed reactor. Hg(0) adsorption tests were conducted at 50 degrees C, under 0.1 and 0.35 ng/cm(3) Hg(0) initial concentrations and with carbon particle sizes ranging between 75-106 and 150-250 microm. Based on the experimental breakthrough data, kinetic studies were performed to investigate the mechanism of adsorption and the rate controlling steps. Kinetic models evaluated include the Fick's intraparticle diffusion equation, the pseudo-first order model, the pseudo-second order model and Elovich kinetic equation. The obtained experimental results revealed that the increase in particle size resulted in significant decrease of breakthrough time and mercury adsorptive capacity, due to the enhanced internal diffusion limitations and smaller external mass transfer coefficients. Additionally, higher initial mercury concentrations resulted in increased breakthrough time and mercury uptake. From the kinetic studies results it was observed that all the examined models describes efficiently Hg(0) breakthrough curves, from breakpoint up to equilibrium time. The most accurate prediction of the experimental data was achieved by second order model, indicating that the chemisorption rate seems to be the controlling step in the procedure. However, the successful attempt to describe mercury uptake with Fick's diffusion model and the first order kinetic model, reveals that the adsorption mechanism studied was complex and followed both surface adsorption and particle diffusion.     

Wavelet penalized likelihood estimation in generalized functional models

The paper deals with generalized functional regression. The aim is to estimate the influence of covariates on observations, drawn from an exponential distribution. The link considered has a semiparametric expression: if we are interested in a functional influence of some covariates, we authorize others to be modeled linearly. We thus consider a generalized partially linear regression model with unknown regression coefficients and an unknown nonparametric function. We present a maximum penalized likelihood procedure to estimate the components of the model introducing penalty based wavelet estimators. Asymptotic rates of the estimates of both the parametric and the nonparametric part of the model are given and quasi-minimax optimality is obtained under usual conditions in literature. We establish in particular that the ? ¹-penalty leads to an adaptive estimation with respect to the regularity of the estimated function. An algorithm based on backfitting and Fisher-scoring is also proposed for implementation. Simulations are used to illustrate the finite sample behavior, including a comparison with kernel- and spline-based methods.     

Distribution of 7Be, 210Pb and 137Cs in watersheds of different scales in the Seine River basin: inventories and residence times

The activity of environmental radionuclides ((7)Be, (210)Pb and (137)Cs) was monitored in nested catchments, inside the Seine River basin. Suspended matter data was collected at 8 different watersheds, ranging from order 1 to order 7, and ranging in size over 4 orders of magnitude. Suspended matter was analyzed for (210)Pb, (137)Cs and (7)Be, and used to calculate the flux of sediments out of each watershed. Monthly atmospheric flux data of (210)Pb and (7)Be was analyzed to assess the input flux of each into the watersheds, taking into account the rainfall during sampling periods. Taking advantage of the different half-lives of (7)Be (53 days) and (210)Pb (22 years), a two-box model was built for each of the catchments following a methodology previously developed by Dominik et al. [Dominik J, Burrus D, Vernet JP. Transport of the environmental radionuclides in alpine watershed. Earth Planet Sci Letters 1987; 84: 165-180.]. The model divides the watershed into a soil box and a rapid reservoir and provides insight into the removal rate of suspended matter from the surrounding watershed. The model enables the assessment of the surface area and the residence time of slow and rapid reservoirs to describe the fate of contaminants of atmospheric origin inside the river basin. The model was improved by considering the dissolved fraction in the total flux and adding the (137)Cs inventory as an additional constraint. The effects of these changes are discussed. Residence times in the soil box, characterized by low transport velocity, range between 4800 years at Melarchez (order 1) to about 30000 years at Andresy and Poses (order 7). They remain constant in each watershed over a large range of variation of atmospheric fluxes of (7)Be and (210)Pb during the whole study, but are sensitive to SM variations. The residence time in the rapid box, which includes the surface of the river and immediate surroundings, is less than one year, while its surface area is in the range 0.6% to 2.2% of the total catchment area. They are sensitive to (7)Be atmospheric flux variations. The two-box model was used to estimate the amount of the radionuclides in each reservoir. Inventories appear to be constant from one watershed to the next. The (7)Be inventory ratio in the rapid and slow boxes expresses the rate of particle-reactive atmospheric pollutants that will be rapidly delivered to the river.