11 research outputs found
Improving Estimation in Speckled Imagery
No full textthis paper is to compare the performance, in terms of bias, of four di#erent estimators for the homogeneity parameter of a distribution that proves to be very useful in modelling image data contaminated by speckle noise, as SAR (Synthetic Aperture Radar), sonar, ultrasound-B and laser imagery. Those data are generated by systems that employ coherent illumination, which generate a stochastic multiplicative noise along with the information. The statistical properties of this noise have been extensively studied in the last thirty years (see Goodman 1976, Oliver & Quegan 1998). Most of the proposed models for image data can be considered as variations of a general framework, the multiplicative model. The multiplicative model assumes the resulting image to be the product of two statistically independent random variables which are related, respectively, to the target information and the speckle noise. Within this framework, a particularly sucessful distribution that was proposed as a model for data coming from images of extremely heterogeneous regions (e.g., urban areas in SAR imagery) is that one called the I distribution. This distribution was introduced by Frery, Muller, Yanasse & Sant'Anna (1997a) and practical applications have shown its outstanding performance in fitting that type of data (for an application to statistical classification see Mejail, Jacobo-Berlles, Frery & Bustos 2003). Recent results propose it as a universal model for speckled data (see Mejail, Frery, Jacobo-Berlles & Bustos 2001). Robust procedures have been also studied for particular cases of this law (see Frery, Sant'Anna, Mascarenhas & Bustos 1997b, Bustos, Lucini & Frery 2002), and improved inference using resampling is presented in Cribari-Neto, Frery & Silva (2002
The Polarimetric G Distribution for SAR Data Analysis
No full textRemote sensing data, and radar data in particular, have become an essential tool for enviromental studies. Many airborne polarimetric sensors are currently operational, and many more will be available in the near future including spaceborne platforms. The signal-to-noise ratio of this kind of imagery is lower than that of optical information requiring, thus, a careful statistical modelling. This modelling may lead to useful or useless techniques for image processing and analysis, according to the agreement between the data and their assumed properties. Several distributions have been used to describe Synthetic Aperture Radar (SAR) data. Many of these univaritate laws arise by assuming the multiplicative model, such as Rayleigh, Square Root of Gamma, Exponential, Gamma, and the class of the I distributions. The adequacy of these distributions depends on the detection (amplitude, intensity, complex etc.), the number of looks, and the homogeneity of the data. In Frery, Muller, Yanasse and Sant'Anna (1997) another class of univariate distributions, called G, was proposed to model extremely heterogeneous clutter, such as urban areas, as well as other types of clutter. This paper extends the univariate family to the multivariate multilook polarimetric situation: the P law. The new family has the classical polarimetric multilook P distribution as a particular case, but another special case is shown more flexible and tractable, while having the same number of parameters and fully retaining their interpretability: the P law. The main properties of this new multivariate distribution are shown. Some results of modelling polarimetric data using P distribution are presented for two airborne polarimetric systems and a variety of targets, showing its expresiveness b..
UJM at CLEF in Author Verification based on optimized classification trees
No full texthttp://ceur-ws.org/Vol-1180/CLEF2014wn-Pan-FreryEt2014.pdfInternational audienceThis article describes our proposal for the Author Identification task in the PAN CLEF Challenge 2014. We have adopted a machine learning ap- proach based on several representations of the texts and on optimized decision trees which have as entry various attributes and which are learned for every train- ing corpus separately for this classification task. Our method ranked us at the 2nd place with an overall AUC of 70.7%, and C@1 of 68.4% and, between the 1st and the 6th place on the six corpora
UJM at CLEF in Author Verification based on optimized classification trees
No full texthttp://ceur-ws.org/Vol-1180/CLEF2014wn-Pan-FreryEt2014.pdfInternational audienceThis article describes our proposal for the Author Identification task in the PAN CLEF Challenge 2014. We have adopted a machine learning ap- proach based on several representations of the texts and on optimized decision trees which have as entry various attributes and which are learned for every train- ing corpus separately for this classification task. Our method ranked us at the 2nd place with an overall AUC of 70.7%, and C@1 of 68.4% and, between the 1st and the 6th place on the six corpora
Improved Estimation of Clutter Properties in Speckled Imagery
No full textThis paper's aim is toeval#9J# the e#ectiveness of bootstrap methods in improving estimation ofcl####q properties in speckl#dREJ5#'dl Estimation is performedby standardmaximum l#rdmaximumdWWW# s. We show that estimators obtainedthis way can be quite biasedin #nite sampl#dW anddevel## bias correction schemes using bootstrapresampl#dx# Inparticul#Eq we propose a bootstrapping scheme which is an adaptation of that proposed by Efron (J. Amer. Statist. Assoc. 85 (1990) 79). The proposedbootstrap does not require the quantity of interest to have cl#ved form, as does Efron'soriginal proposal# The adaptation we suggest isparticul#EEx important since the maximum l#imumd#599JdRE#ExW of interest does not have acl##J'dRE#E We show that thisparticul#d bootstrapping scheme outperforms al#performs forms of bias reduction mechanisms, thusdel##q'#9d more accurate inference. Weal#E considerinterval estimation using bootstrap methods, andshow that aparticul#J parametric bootstrap-basedcon#denceinterval is typical#9 morerel#d#x9 than both the asymptotic con#denceinterval and other bootstrap-based con#dence interval#E Anappl###dR#q toreal data is presented and discussed. c 2002El#2dW## Science B.V.Al# rights reserved
Image Formation in Vibro-acoustography with Sector Array Transducers
No full textThis paper presents the image formation process in vibro-acoustography for systems based on sector array transducers. These transducers are an alternative to annular concave transducers. They represent an innovative technique deserving detailed assessment in vibro-acoustography applications. The system point-spread function (PSF) is defined in terms of the acoustic emission of a point-target in response to the dynamic ultrasound radiation force. This force is produced by two overlapping ultrasound beams. We calculate the radiation force on the target in a nonviscous fluid using the plane wave approximation for the ultrasound beams. The beamforming of sector array transducers is analyzed through linear acoustics. An expression for the velocity potential produced by sector arrays is derived, and the vibro-acoustography PSF is evaluated numerically. An experimental setup is design to validate the theory; the comparison is made using location and amplitude of sidelobes and spatial resolution defined by the PSF. Results show that the computed PSF is in full agreement with the PSF obtained experimentally. 1
Simulation and assessment of flooding with desktop virtual reality
No full textAbstract. A system that builds desktop virtual reality models based on topographic maps and its application to tidal dynamics analysis is shown. The virtual reality model allows the user to explore the scene from any possible point of view, also permitting to alter the level of the sea simulating tides and flooding. Flooded and dry areas are visible at each water level, and the flooding pattern can be assessed. The system allows the choice of texture and colors for the elevation model and the creation and edition of routes. The application receives as input a digital elevation model, and builds a virtual reality description using Virtual Reality Modeling Language, which allows good interaction between user and model. A case study is presented with remote sensing data acquired over the USA
A benchmark for despeckling filters
No full textThe validation of filters for SAR image denoising is mostly performed through personal choices of experimentation on a set of author-selected synthetic and true SAR images, which makes the objective elucidation of their benefits over well-established filtering methods hard. Following the example of other scientific fields, this situation can be alleviated by providing a set of reference images to test new methods on thorough quantitative measures of quality. We propose an extensible benchmark for speckle denoising techniques. We also stress the need to establish a repository with a pool of error-free and correctly codified denoising filters and quality estimators.454451
In Virtual Reality 5:1-10. 2001 A Synthetic Actor Model for Long-Term Computer
No full textAbstract. Virtual Reality and Artificial Intelligence provide suitable techniques to improve computer games quality. While the former offers mechanisms to model environment and characters physical features, the latter provides models and tools for building characters, namely Synthetic Actors or Believable Agents, which can exhibit intelligent social behavior and express personality and emotions. The current architecture proposals for Synthetic Actors do not fully meet the requirements for long-term games development. In long-term games, such as strategy and adventure ones, it is necessary to guarantee both personality stability and reactive emotional responses, which may be contradictory. In this work, we propose a new Synthetic Actor model that tightly connects emotions and social attitudes to personality, providing a longterm coherent behavior. This model has been applied to two games presented here as case studies
Vitamin B12 and folate status during pregnancy among Saudi population
No full textT2DM is a growing health problem worldwide. It is now increasingly being diagnosed
earlier in life. The factors involved in such an epidemic are complex. The intrauterine
environment has long been known as an important contributor to many diseases
including metabolic disorders such as T2DM. Recently, there is emerging evidence for
maternal micronutrients affecting vital developmental processes in utero which can
adversely “programme” the offspring to develop metabolic disorders in later life. Thus,
“gene-diet” interaction during foetal development is likely to be a significant
contributor to the epidemic of T2DM. In particular, the intrauterine imbalance between
the two related vitamins, vitamin B12 and folate, affect DNA methylation and in turn
programme the foetus for the whole life. Evidence from mandatory folic acid
fortification studies suggests that in the presence of adequate folate, neural tube defects
due to vitamin B12 insufficiency have tripled. In India, children born to mothers with
“high folate and low vitamin B12” had higher adiposity and insulin resistance.
Therefore, micronutrient status during pregnancy is likely to have a significant impact
on the metabolic risk of the offspring. This thesis examines whether vitamin B12
insufficiency is prevalent in pregnancy, especially in a non-vegetarian population across
the world as well as the Saudi pregnant population. As estimated intake is an accepted
measure for micronutrient levels, we also examined the relationship between estimated
vitamin B12 and folate intake with actual levels in the blood. We have found that
vitamin B12 insufficiency was not uncommon during pregnancy across the world even
in the non-vegetarian population and is also common in the Saudi population.
Surprisingly, vitamin B12 insufficiency was observed in 50% of the tested population
even in the presence of adequate vitamin B12 intake. In addition, we have also shown
for the first time in the Saudi population that maternal BMI is inversely related to vitamin B12 levels, particularly in pregnancy. Even though we have shown a similar (or
worse) picture in mothers with gestational diabetes, this study needs to be replicated, as
our numbers are too small. Prospective studies linking the role of vitamin B12
insufficiency especially in the presence of high folate on birth outcomes in the Saudi
population as well as intervention studies investigating the role of vitamin B12
supplementation in women of childbearing age and in pregnancy are urgently needed
