Navegando por Data de Publicação, começando com "2009-07-24"

Filtrar resultados por ano ou mês

Agora exibindo 1 - 3 de 3

Combinação de múltiplos classificadores para reconhecimento de face humana
(Universidade Federal de São Carlos, 2009-07-24) Salvadeo, Denis Henrique Pinheiro; Mascarenhas, Nelson Delfino d'Ávila; https://lattes.cnpq.br/0557976975338451; https://lattes.cnpq.br/1475921082905793
Lately, the human face object has been exploited by the advent of systems involving biometrics, especially for applications in security. One of the most challenging applications is the problem of human face recognition, which consists of determining the correspondence between an input face and an individual from a database of known persons. The process of face recognition consists of two steps: feature extraction and classification. In the literature of face recognition, different techniques have been used, and they can be divided into holistic techniques (implicit feature extraction), feature-based techniques (explicit feature extraction) and hybrid techniques (involving the two previous). In many articles, holistic techniques have proved to be most efficient and generally they involve methods of statistical pattern recognition as Principal Component Analysis (PCA), Independent Component Analysis (ICA), Linear Discriminant Analysis (LDA), Support Vector Machines (SVM), and Neural Networks. For problems such as human face recognition in digital images, a crucial point is the ability to generalize. The solution for this problem is complex due to the high dimensionality of data and the small number of samples per person. Using a single classifier would reduce the ability of recognition, since it is difficult to design a single classifier in these conditions that capture all variability that span the human faces spaces. Thus, this work proposes to investigate the combination of multiple classifiers applied to the problem of face recognition, defining a new scheme to resolve this problem, varying the feature extraction with PCA and some its variants and LDA, K-Nearest Neighbor (K-NN) and Maximum Likelihood (MaxLike) classifiers and several trainable or not trainable methods for combining classifiers. Still, to mitigate the problem of small sample size (SSS), a technique for regularizing the covariance matrix was used. Finally, to assess the classification performance, Holdout and Resubstitution methods were used to partition the data set and the Kappa coefficient and Z and T statistics were used to measure the performance of the proposed scheme. From the experiments it was concluded that the best sub-schemes were the RBPCA/MaxLike-PCA/NN-KL5/NN classifiers combined by the Majorite Vote Rule for the ORL database and the RLDA+RPCA/MaxLike-KL4/NNKL5/ NN classifiers combined by the Sum Rule for the AR database, obtaining Kappa coefficients of 0.956 (mean) and 0.839, respectively. Besides that, it has been determined that these sub-schemes are robust to pose (ORL database), illumination and small change of the facial expression, but they were affected by occlusions (AR database).
Desenvolvimento de estratégias para a determinação espectrofotométrica sequencial em fluxo de Co (II) e Mn (II)
(Universidade Federal de São Carlos, 2009-07-24) Ferreira, Juliana Aparecida; Pereira Filho, Edenir Rodrigues; https://lattes.cnpq.br/3394181280355442; https://lattes.cnpq.br/9417279435389816
In this study a kinetic-spectrophotometric method was developed for sequential determination of Co and Mn in pharmaceutical samples. The method was based on the catalytic effect of the analytes in the oxidation reaction of Tiron by H2O2 in basic medium. In the case of Mn it was employed the activator reagent 2,2-bipyridine to effective catalysis. The FIA system was projected taking into account the catalytic properties of each analyte in the indicator reaction. Two factorial designs were applied, initially, a fractionary factorial design 29-5 followed by a complete factorial design 23 to select the important variables, and then a factorial design 22 + central point + star for optimizing of FIA system. Fractionary factorial designs 27-2 were applied in the study of possible interferents for Co and Mn. The main interferents were mathematically described using factorial + central point + star designs. The pharmaceutical samples were digested by concentrated HNO3 in digestor block. To evaluate the accuracy of the developed method, samples (n=3) were analyzed by FS-FAAS and based on the determination concentrations, these were diluted to posterior analysis aplyingby the developed method. Limits of detection (LOD) and quantification (LOQ) for FS-FAAS were 3.9 and 13.2 μg L-1 for Co; 110.0 and 366.0 μg L-1 for Mn, respectively. To developed method the LOD and LOQ were 0.055 and 0.18 μg L-1 for Co, 9.76 and 32.5 μg L-1 for Mn, respectively. Comparison between the two methods was evaluated by paired t-test. At 99% confidence level, the values did not differ statistically for Co. However, for Mn, the concentration values agreed but this data cannot be statistically confirmed due to differences of standard deviations of the two methods. The repeatability was 3.2% for Co and 8.0% for Mn. The sampling frequency was 25 samples h-1.
Análise comparativa da variação genética entre os estoques cultivado e natural de Prochilodus argenteus: implicações para o repovoamento de rios
(Universidade Federal de São Carlos, 2009-07-24) Campos, Wagner Narciso de; Hatanaka, Terumi; https://lattes.cnpq.br/1666878061867996; https://lattes.cnpq.br/3793013870234455
Introduction of hatchery fish in river has been the major using strategy for river fishery rehabilitation, working to attenuate river damage in short time. But there are effective risks about their efficiency and results, concerning the pool gene preservation and illness proliferation, beyond others ecologies and economics aspects. At São Francisco river, during two decades it was used this kind of procedure with some species of fish, but there are no quantitative data about their efficiency. Prochilodus argenteus, popular known as Curimatápacu, is pointed as the most important economic specie at São Francisco basin, responsible for 50% of all fishery production. Therefore, this project goals to trace the genetic profile of a Prochilodus argenteus hatchery population, looking for identify stock genetic variability through 13 SSR loci previous described, and to compare theses data with others described in literature from wild stocks. Tests made with Microchecker, Genepop, and Fsat softwares detected low homozygosis; reduced allelic variability when compared with wild population; increase in heterozygosis when compared with expected; and negative inbreed coefficient, these data got a trace for bottleneck phenomena in a cultured population, and Bottleneck software tests confirm this hypothesis. When the same test was ruled with wild population data results very far from those found in hatchery populations, give a high reliability. To conclude, at cultured and/or management programs are necessary the knowledge of genetic profile population (hatchery or wild) to prevent damaging introgression in wild population and consequently extinction. These data can be used to made fishery river rehabilitation program more efficient, and so, adding with information to fish procreation, industrial fishery and biologic fish conservation studies.