Algoritmo SOM com estrutura hierárquica e dinâmica aplicado a compressão de imagens

ln this work the implementation of the SOM (Self Organizing Maps) algorithm or Kohonen neural network is presented in the form of hierarchical structures, applied to the compression of images. The main objective of this approach is to develop an Hierarchical SOM algorithm with static structure and another one with dynamic structure to generate codebooks (books of codes) in the process of the image Vector Quantization (VQ), reducing the time of processing and obtaining a good rate of compression of images with a minimum degradation of the quality in relation to the original image. Both self-organizing neural networks developed here, were denominated HSOM, for static case, and DHSOM, for the dynamic case. ln the first form, the hierarchical structure is previously defined and in the later this structure grows in an automatic way in agreement with heuristic rules that explore the data of the training group without use of external parameters. For the network, the heuristic mIes determine the dynamics of growth, the pruning of ramifications criteria, the flexibility and the size of children maps. The LBO (Linde-Buzo-Oray) algorithm or K-means, one ofthe more used algorithms to develop codebook for Vector Quantization, was used together with the algorithm of Kohonen in its basic form, that is, not hierarchical, as a reference to compare the performance of the algorithms here proposed. A performance analysis between the two hierarchical structures is also accomplished in this work. The efficiency of the proposed processing is verified by the reduction in the complexity computational compared to the traditional algorithms, as well as, through the quantitative analysis of the images reconstructed in function of the parameters: (PSNR) peak signal-to-noise ratio and (MSE) medium squared error

Técnicas visuais de localização e mapeamento simultâneos sem extração de primitivas geométricas da imagem

In Simultaneous Localization and Mapping (SLAM - Simultaneous Localization and Mapping), a robot placed in an unknown location in any environment must be able to create a perspective of this environment (a map) and is situated in the same simultaneously, using only information captured by the robot s sensors and control signals known. Recently, driven by the advance of computing power, work in this area have proposed to use video camera as a sensor and it came so Visual SLAM. This has several approaches and the vast majority of them work basically extracting features of the environment, calculating the necessary correspondence and through these estimate the required parameters. This work presented a monocular visual SLAM system that uses direct image registration to calculate the image reprojection error and optimization methods that minimize this error and thus obtain the parameters for the robot pose and map of the environment directly from the pixels of the images. Thus the steps of extracting and matching features are not needed, enabling our system works well in environments where traditional approaches have difficulty. Moreover, when addressing the problem of SLAM as proposed in this work we avoid a very common problem in traditional approaches, known as error propagation. Worrying about the high computational cost of this approach have been tested several types of optimization methods in order to find a good balance between good estimates and processing time. The results presented in this work show the success of this system in different environments

Estudo avaliativo de um algoritmo genético auto-organizável e multiobjetivo utilizando aprendizado de máquina para aplicações de telecomunicações

This paper presents an evaluative study about the effects of using a machine learning technique on the main features of a self-organizing and multiobjective genetic algorithm (GA). A typical GA can be seen as a search technique which is usually applied in problems involving no polynomial complexity. Originally, these algorithms were designed to create methods that seek acceptable solutions to problems where the global optimum is inaccessible or difficult to obtain. At first, the GAs considered only one evaluation function and a single objective optimization. Today, however, implementations that consider several optimization objectives simultaneously (multiobjective algorithms) are common, besides allowing the change of many components of the algorithm dynamically (self-organizing algorithms). At the same time, they are also common combinations of GAs with machine learning techniques to improve some of its characteristics of performance and use. In this work, a GA with a machine learning technique was analyzed and applied in a antenna design. We used a variant of bicubic interpolation technique, called 2D Spline, as machine learning technique to estimate the behavior of a dynamic fitness function, based on the knowledge obtained from a set of laboratory experiments. This fitness function is also called evaluation function and, it is responsible for determining the fitness degree of a candidate solution (individual), in relation to others in the same population. The algorithm can be applied in many areas, including in the field of telecommunications, as projects of antennas and frequency selective surfaces. In this particular work, the presented algorithm was developed to optimize the design of a microstrip antenna, usually used in wireless communication systems for application in Ultra-Wideband (UWB). The algorithm allowed the optimization of two variables of geometry antenna - the length (Ls) and width (Ws) a slit in the ground plane with respect to three objectives: radiated signal bandwidth, return loss and central frequency deviation. These two dimensions (Ws and Ls) are used as variables in three different interpolation functions, one Spline for each optimization objective, to compose a multiobjective and aggregate fitness function. The final result proposed by the algorithm was compared with the simulation program result and the measured result of a physical prototype of the antenna built in the laboratory. In the present study, the algorithm was analyzed with respect to their success degree in relation to four important characteristics of a self-organizing multiobjective GA: performance, flexibility, scalability and accuracy. At the end of the study, it was observed a time increase in algorithm execution in comparison to a common GA, due to the time required for the machine learning process. On the plus side, we notice a sensitive gain with respect to flexibility and accuracy of results, and a prosperous path that indicates directions to the algorithm to allow the optimization problems with "η" variables

Contribuições a técnicas de agrupamento e visualização de dados multivariados utilizando mapas auto-organizáveis

Self-organizing maps (SOM) are artificial neural networks widely used in the data mining field, mainly because they constitute a dimensionality reduction technique given the fixed grid of neurons associated with the network. In order to properly the partition and visualize the SOM network, the various methods available in the literature must be applied in a post-processing stage, that consists of inferring, through its neurons, relevant characteristics of the data set. In general, such processing applied to the network neurons, instead of the entire database, reduces the computational costs due to vector quantization. This work proposes a post-processing of the SOM neurons in the input and output spaces, combining visualization techniques with algorithms based on gravitational forces and the search for the shortest path with the greatest reward. Such methods take into account the connection strength between neighbouring neurons and characteristics of pattern density and distances among neurons, both associated with the position that the neurons occupy in the data space after training the network. Thus, the goal consists of defining more clearly the arrangement of the clusters present in the data. Experiments were carried out so as to evaluate the proposed methods using various artificially generated data sets, as well as real world data sets. The results obtained were compared with those from a number of well-known methods existent in the literature

Controle feedback de nível baseado em sensor de imagem aplicado ao equipamento misturador-decantador à inversão de fases (MDIF®)

The treatment of wastewaters contaminated with oil is of great practical interest and it is fundamental in environmental issues. A relevant process, which has been studied on continuous treatment of contaminated water with oil, is the equipment denominated MDIF® (a mixer-settler based on phase inversion). An important variable during the operation of MDIF® is the water-solvent interface level in the separation section. The control of this level is essential both to avoid the dragging of the solvent during the water removal and improve the extraction efficiency of the oil by the solvent. The measurement of oil-water interface level (in line) is still a hard task. There are few sensors able to measure oil-water interface level in a reliable way. In the case of lab scale systems, there are no interface sensors with compatible dimensions. The objective of this work was to implement a level control system to the organic solvent/water interface level on the equipment MDIF®. The detection of the interface level is based on the acquisition and treatment of images obtained dynamically through a standard camera (webcam). The control strategy was developed to operate in feedback mode, where the level measure obtained by image detection is compared to the desired level and an action is taken on a control valve according to an implemented PID law. A control and data acquisition program was developed in Fortran to accomplish the following tasks: image acquisition; water-solvent interface identification; to perform decisions and send control signals; and to record data in files. Some experimental runs in open-loop were carried out using the MDIF® and random pulse disturbances were applied on the input variable (water outlet flow). The responses of interface level permitted the process identification by transfer models. From these models, the parameters for a PID controller were tuned by direct synthesis and tests in closed-loop were performed. Preliminary results for the feedback loop demonstrated that the sensor and the control strategy developed in this work were suitable for the control of organic solvent-water interface level

O auto da morte e da vida : a escrita barroca de João Cabral de Melo Neto

This dissertation, entitled O Auto da Morte e da Vida: A escrita barroca de João Cabral de Melo Neto, has the aim of analising, interpreting, in a baroque perspective, Cabral s writing in the poem/play Morte e vida severina Auto de Natal Pernambucano, taking as basis the theories of Eugênio D´Ors, Severo Sarduy, Omar Calabrase, Lezama Lima, Afonso Ávila, Affonso Romano de Sant´Anna and others cited in the body of this work. During the analisys we feature confluences, relations, similarities, identification between the Baroque of the counter reformation and the modern Baroque or Neobaroque. We seek to comprehend the baroque which is new in the XX century and Cabral s poetry as an element of the contemporaneity, by updating the concept of the Baroque in the 1600s, when it is detected in its purest characteristic in human relation (the life of the Northwestern brazilian) through an intangible reality (the death). The Baroque as a cultural summary of a period of instability and transformation, with the power of dismantling an already established poetry. The fight between words and things, language and reality

A construção da imagem de leitor em memórias de leitura

In this work, we analyzed reading memories of mother language teachers in continuing education context. Our objective was to understand how each individual has built his/her reader image. Our theoretical approach to the construction of selfimage was based on the concept of discursive ethos, understanding it with Charaudeau (2006) as something constructed in the intersection of glances (of the self and the other). To understand how each teacher has built his/her reader image in that intertwining of glances (of the self and the other) we are on the contributions of Bakhtin (2003, 2010b) on exotopic glance or distant/external glance. Therefore, in the analysis, we tried to capture the exotopic glance of the teachers about themselves in the various stages of their reader formation and from our exotopic look of researcher; we gave provisional finish of the reader image that teachers built of themselves. For the analysis, we adopted other theoretical assumptions: about genres, theme, composition and style, statement and social voices we based on Bakhtin (1997, 2003, 2010a, 2010b); on the notion of the discursive ethos we anchored in studies conducted by Maingueneau (2008a, 2008b); about reading, we adopted the theoretical references of Rojo (2005, 2008, 2009a, 2009b, 2009c, 2009d), Garcez (2002), Freire (2008) and Silva Neto (2007). For the discursive genre reading memories makes reference to the theme memory as well as is related to the context of teacher training, the study was supported in Aragão (1992) and Nóvoa`s (2007) theory. Situated in the area of Applied Linguistics, the research aligns with qualitative-interpretative approach of socio-historical basis. Finally, from the analysis of the corpus, data that emerged from the findings, we conclude by stating that readers have created images of themselves as active readers, readers interested in both readings, the ones respected and the ones unappreciated by the official culture

Das relações entre imaginário e poesia na obra de Orides Fontela: o regime diurno da imagem

The present work seeks essentially to demonstrate how some anthropological structures of the imaginary, theorized by the French thinker Gilbert DURAND, projected themselves in a noticeable manner in the poetic work of the Brazilian poet Orides FONTELA. Chiefly we will demonstrate with great care in what way this projection and vivification occur through an imagination that not only materializes some fundamental archetypes of the human imagination, but also seeks to organize them through original poems in its form of presenting the poetic discourse, giving here a different contribution of the mythical or religious discourse, privileged places, since always, where the symbolic functions also manifest themselves

O Aleph e seus duplos: mimesis e auto-reflexividade na obra de Jorge Luis Borges

This study approach the Jorge Luis Borges s prose of fiction under the perspective of mimesis and the self-reflexivity. The hypothesis is that the Aleph is a central symbol of the Borges s fictional universe. The rewriting and the retake of this symbol along of his work entail to a reflection about the possibilities and the limits of mimesis. This study is divided in three parts which contain two chapters. The first part Bibliographic revision and conceptual fundaments of inquiry discuss the critical fortune of author (Chapter 1) and the concepts that will give sustentation to the inquiry (Chapter 2). The second part About the Borges s aesthetic project sketch out the literary project defended by Borges that is his conception of the literature and his ideological matrix (Chapter 3) beside his anti-psychologism and his nostalgia of epos (Chapter 4). The third and last part is entitled The Aleph and his doubles. In the chapter 5 this study analyses the short story El Aleph and consider its centrality on the Borges s work. The argument that is on this short story Borges elaborates a reflection about mimesis. In the chapter 6, on the same hand, four short stories will be analysed: Funes el memorioso ; El Libro de Arena ; El evangelio según Marcos and Del rigor en la ciencia . The conclusion that is the Borges s literature is self-awake of its process as such demonstrate its parodic sense and its bookish origin. Hence, the Borges s literature overlapping the mimetic crisis of language and challenge the limits between fiction and reality. However, it doesn t surrender to the nihilist perspective that is closing of literature to the world

Jornalismo cultural em texto, imagem e cor

O tema deste trabalho é o Jornalismo Cultural. Uma apresentação e discussão de temas pertinentes ao jornalismo especializado em cultura nas páginas do jornal diário. A partir da leitura de alguns jornais, se traça considerações gerais acerca dos caminhos do jornalismo e dos estudos sobre jornalismo com o objetivo de apresentar o jornal que hoje se produz no Brasil, em texto, imagem e cor. O resultado é a descoberta do jornal como forma, desde o texto, escolhas lexicais e sintáticas, passando pelas imagens, composição, escolha e determinação da pauta, ao arranjo destes elementos na página, em respeito ao projeto editorial de diagramação. A conclusão a que se chega que há vários "eus" que governam o jornal diário, e que fazem o jornalismo cultural inovador, porta de entrada às mudanças no jornalismo.

A imagem do Brasil na mídia impressa portuguesa: um estudo do caso diário de notícias e público

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior

Modelos auto-similares para rádio fontes extragalácticas.

This work is a detailed study of self-similar models for the expansion of extragalactic radio sources. A review is made of the definitions of AGN, the unified model is discussed and the main characteristics of double radio sources are examined. Three classification schemes are outlined and the self-similar models found in the literature are studied in detail. A self-similar model is proposed that represents a generalization of the models found in the literature. In this model, the area of the head of the jet varies with the size of the jet with a power law with an exponent γ. The atmosphere has a variable density that may or may not be spherically symmetric and it is taken into account the time variation of the cinematic luminosity of the jet according to a power law with an exponent h. It is possible to show that models Type I, II and III are particular cases of the general model and one also discusses the evolution of the sources radio luminosity. One compares the evolutionary curves of the general model with the particular cases and with the observational data in a P-D diagram. The results show that the model allows a better agreement with the observations depending on the appropriate choice of the model parameters.

Comparação e mobilidade, equilíbrio corporal e desempenho muscular segundo a auto-eficácia para quedas em idosas

The aging process modifies various systems in the body, leading to changes in mobility, balance and muscle strength. This can cause a drop in the elderly, or not changing the perceived self-efficacy in preventing falls. Objective: To compare the mobility, body balance and muscle performance according to self-efficacy for falls in community-dwelling elderly. Methods: A cross-sectional comparative study with 63 older (65-80 years) community. Were evaluated for identification data and sociodemographic, cognitive screening using the Mini Mental State Examination (MMSE), effective for the fall of Falls Efficacy Scale International Brazil (FES-I-BRAZIL), Mobility through the Timed Up and Go Test , the balance Berg Balance Scale (BBS) and the Modified Clinical Test tests of Sensory Interaction on Balance (mCTSIB), tandem walk (TW) and Sit to Stand (STS) of the Balance Master® System. Finally, muscle performance by using isokinetic dynamometry. Statistical analysis was performed Student t test for comparison between groups, with p value ≤ 0.05. Results: Comparing the elderly with low-efficacy for falls with high-efficacy for falls, we found significant differences only for the variable Timed Up and Go Test (p = 0.04). With regard to data on balance tests were significant differences in the speed of oscillation firm surface eyes open modified Clinical Test of Sensory Interaction on Test of Balance (p = 0.01). Variables to isokinetic dynamometry were no significant differences in movement knee extension, as regards the variables peak torque (p = 0.04) and power (p = 0.03). Conclusion: The results suggest that, compared to older community with low-and high-efficacy for falls, we observed differences in variables related to mobility, balance and muscle function

Aspectos epidemiológicos, cognitivo-comportamentais e neurofisiológicos do sonho lúcido

Lucid dreaming (LD) is a mental state in which the subject is aware of being dreaming while dreaming. The prevalence of LD among Europeans, North Americans and Asians is quite variable (between 26 and 92%) (Stepansky et al., 1998; Schredl & Erlacher, 2011; Yu, 2008); in Latin Americans it is yet to be investigated. Furthermore, the neural bases of LD remain controversial. Different studies have observed that LD presents power increases in the alpha frequency band (Tyson et al., 1984), in beta oscillations recorded from the parietal cortex (Holzinger et al., 2006) and in gamma rhythm recorded from the frontal cortex (Voss et al., 2009), in comparison with non-lucid dreaming. In this thesis we report epidemiological and neurophysiological investigations of LD. To investigate the epidemiology of LD (Study 1), we developed an online questionnaire about dreams that was answered by 3,427 volunteers. In this sample, 56% were women, 24% were men and 20% did not inform their gender (the median age was 25 years). A total of 76.5% of the subjects reported recalling dreams at least once a week, and about two-thirds of them reported dreaming always in the first person, i.e. when the dreamer observes the dream from within itself, not as another dream character. Dream reports typically depicted actions (93.3%), known people (92.9%), sounds/voices (78.5%), and colored images (76.3%). The oneiric content was related to plans for upcoming days (37.8%), and memories of the previous day (13.8%). Nightmares were characterized by general anxiety/fear (65.5%), feeling of being chased (48.5%), and non-painful unpleasant sensations (47.6%). With regard to LD, 77.2% of the subjects reported having experienced LD at least once in their lifetime (44.9% reported up to 10 episodes ever). LD frequency was weakly correlated with dream recall frequency (r = 0.20, p <0.001) and was higher in men (χ2=10.2, p=0.001). The control of LD was rare (29.7%) and inversely correlated with LD duration (r=-0.38, p <0.001), which is usually short: to 48.5% of the subjects, LD takes less than 1 minute. LD occurrence is mainly associated with having sleep without a fixed time to wake up (38.3%), which increases the chance of having REM sleep (REMS). LD is also associated with stress (30.1%), which increases REMS transitions into wakefulness. Overall, the data suggest that dreams and nightmares can be evolutionarily understood as a simulation of the common situations that happen in life, and that are related to our social, psychological and biological integrity. The results also indicate that LD is a relatively common experience (but not recurrent), often elusive and difficult to control, suggesting that LD is an incomplete stationary stage (or phase transition) between REMS and wake state. Moreover, despite the variability of LD prevalence among North Americans, Europeans and Asians, our data from Latin Americans strengthens the notion that LD is a general phenomenon of the human species. To further investigate the neural bases of LD (Study 2), we performed sleep recordings of 32 non-frequent lucid dreamers (sample 1) and 6 frequent lucid dreamers (sample 2). In sample 1, we applied two cognitive-behavioral techniques to induce LD: presleep LD suggestion (n=8) and light pulses applied during REMS (n=8); in a control group we made no attempt to influence dreaming (n=16). The results indicate that it is quite difficult but still possible to induce LD, since we could induce LD in a single subject, using the suggestion technique. EEG signals from this one subject exhibited alpha (7-14 Hz) bursts prior to LD. These bursts were brief (about 3s), without significant change in muscle tone, and independent of the presence of rapid eye movements. No such bursts were observed in the remaining 31 subjects. In addition, LD exhibited significantly higher occipital alpha and right temporo-parietal gamma (30-50 Hz) power, in comparison with non-lucid REMS. In sample 2, LD presented increased frontal high-gamma (50-100 Hz) power on average, in comparison with non-lucid REMS; however, this was not consistent across all subjects, being a clear phenomenon in just one subject. We also observed that four of these volunteers showed an increase in alpha rhythm power over the occipital region, immediately before or during LD. Altogether, our preliminary results suggest that LD presents neurophysiological characteristics that make it different from both waking and the typical REMS. To the extent that the right temporo-parietal and frontal regions are related to the formation of selfconsciousness and body internal image, we suggest that an increased activity in these regions during sleep may be the neurobiological mechanism underlying LD. The alpha rhythm bursts, as well as the alpha power increase over the occipital region, may represent micro-arousals, which facilitate the contact of the brain during sleep with the external environment, favoring the occurrence of LD. This also strengthens the notion that LD is an intermediary state between sleep and wakefulness

Influência da auto seleção à luz no ritmo circadiano de atividade motora em Callithrix Jacchus

Marmosets, Callithrix jacchus, are strictly diurnal animals. The motor activity rhythmicity is generated by the circadian timing system and is modulated by environmental factors, mainly by photic stimuli that compose the light-dark cycle. Photic stimuli can reset the biological oscilators changing activity motor pattern, by a mechanism called entrainment. Otherwise, light can act directly on expressed rhythm, without act on the biological oscillators, promoting the masking. Thus, photic stimuli can synchronize the circadian activity rhythm (CAR) by two distinct mechanisms, acting isolated or at a combined way. Among the elements that can influence photic synchronization, the duration and time of photic exposure is pointed out. If in the natural environment the marmoset can choose places of different intensity illumination and is synchronized to light-dark cycle (LD), how the photic synchronization mechanism can be evaluated in laboratory by light self-selection? With objective to response this question, four adult male marmosets were studied at two conditions: with and without sleeping box. The animals were submitted to a LD cycle (12:12/ 350:2 lx) and constant light (LL: 350 lx) conditions in individual cages with an opaque sleeping box, that permitted the light self-selection. At the room, the temperature was 25.6 ºC (± 0.3 ºC) and humidity was 78.7 (± 5%). The motor activity was recorded at 5 min bins by infrared movement sensors installed at the top of the cages. The motor activity profile was distinct at the two conditions: without the sleeping box protection against light, the activity frequency was higher at CT 11-12 (ANOVA; F(3.23) = 62.27; p < 0.01). Also, the duration of the active phase (α) was prolonged of about 1 h (t test, p < 0.05) and the animals showed a significant delay on the activity onset and offset (t test, p < 0.05) and at the acrophase (confidence intervals of 5%) of CAR. In LL, the light continuous exposure prolonged the active phase and influenced the endogenous expression of the circadian activity rhythm period. From the result analysis, it is concluded that the light self-selection can modify several parameters of CAR in marmosets, allowing the study of the synchronization mechanism using the burrow model. Thus, without sleeping box there was a phase delay between the CAR and LD (entrainment) and an increase of activity near lights off (positive masking). Furthermore, in LL, the light continuous exposure modifies α and the endogenous expression of CAR. It is suggested that the light self-selection might be take into account at investigations that evaluate the biological rhythmicity in marmosets