Reconhecimento de caracteres alfanuméricos de placas em imagens de veículos

Sistemas de visão artificial são cada vez mais usados para auxiliar seres humanos a realizar diferentes tarefas. Estes sistemas são capazes de reconhecer padrões em imagens complexas. Técnicas de visão computacional têm encontrado crescente aplicação em estudos e sistemas de controle e monitoração de tráfego de automóveis. Uma das áreas de pesquisa que tem sido objeto de estudo por diferentes grupos é a leitura automática de placas de matrículas como forma de detectar transgressores, encontrar carros roubados ou efetuar estudos de origem/destino [BAR99]. Com o constante crescimento do volume de tráfego de automóvel e a limitada capacidade dos sensores convencionais, especialistas da área recorrem a técnicas de identificação automática de veículos para obter dados relativos ao escoamento de tráfego. A identificação automática de veículos tem tido essencialmente duas abordagens distintas: a utilização de transponders e a utilização de técnicas de visão computacional [INI85] . Estas são essencialmente úteis em casos em que não é viável obrigar os motoristas a instalar transponders em seus automóveis. No entanto, essas técnicas são mais sensíveis às condições atmosféricas e de iluminação tais como nevoeiros, chuva intensa, luz noturna, reflexos em superfícies, etc. Este trabalho apresenta um estudo de diversas técnicas de processamento de imagem objetivando o aperfeiçoamento de um sistema de identificação automática de placas de veículos. Este aperfeiçoamento está relacionado com a diminuição do tempo de execução necessário à localização e reconhecimento dos caracteres contidos nas placas dos veículos bem como a melhorar a taxa de sucesso no seu reconhecimento. A primeira versão do sistema de identificação da placas de veículos descrito em [SOU2000], desenvolvido no CPG-EE da UFRGS, denominado SIAV 1.0, localiza e extrai 91,3% das placas corretamente mas apresenta uma taxa de reconhecimento das placas de 37,3%, assim como um tempo de processamento não satisfatório. Neste trabalho, cujo sistema desenvolvido é denominado SIAV 2.0, a imagem é previamente processada através da aplicação de técnicas de realce da imagem. O principal objetivo das técnicas de realce é processar a imagem de modo que o resultado seja mais apropriado para uma aplicação específica do que a imagem original [GON93]. O sistema busca melhorar a qualidade da imagem eliminando ou suavizando sombras e reflexos presentes na cena em virtude da iluminação não controlada. Visando um menor tempo de execução durante o tratamento e análise da imagem um estudo estatístico baseado na distribuição gaussiana foi realizado de maneira a restringir a área de análise a ser processada. O SIAV possui duas redes neurais como ferramentas de reconhecimento de caracteres. A partir da análise dos diferentes modelos de redes neurais empregados na atualidade, foi desenvolvida uma nova arquitetura de rede a ser utilizada pelo SIAV 2.0 que oferece uma taxa de reconhecimento superior a rede neural usada no SIAV 1.0. Visando um melhor tempo de execução, a implementação em hardware dedicado para este modelo é abordado. Os testes foram realizados com três bancos de imagens obtidas por câmeras diferentes, inclusive por dispositivo "pardal" comercial. Estes testes foram realizados para verificar a efetividade dos algoritmos aperfeiçoados.

Processo de escolha de bancos de imagens: aplicações no marketing business to business

The general objective of this research is to verify which attributes are most relevant to a stock photography agency that represent the purchaser's standards of choice. For this objective to be accomplished, qualitative interviews with the customers of stock photography agencies had been made in order to raise the attributes considered with relevance in the process of choice for the Brazilian stock photography agency market. The levels of each attribute to be tested had arisen through the mapping of the products and services offered by the competition and the relative weight assigned to each one of the attributes related to the research in the choices for a stock photography agency. A transversal study was made with a sample of stock photography agencies customers from Brazil¿s Southeastern region. For the analysis of this data, a survey method was used, that, according to MALHOTRA (2006: 182), involves a structuralized questionnaire to be answered by an interview that elucidates specific information, in which the questions had a predetermined order. This questionnaire was presented through interviews on the Internet. The method adopted for this analysis was a joint analysis. A sampling of 1000 customer stock photography agencies were selected, which were represented by a didactic book editor, advertising agencies, editorial companies, etc., in the states of Rio de Janeiro, São Paulo, Espírito Santo and Minas Gerais. They had been obtained by 16 stimulations, or cards, for the application of the companies. Of the sample of 1000 customers of stock photography agencies, 7,8% responded to the research. The attribute most relevant in accordance with the presented results is the availability of images in high resolution for layout. Its relative importance was 40,78%. At the end of the study, the management implications were propitiated to the stock photography agencies managers inabling them to be more competitive.

Medo e violência na cidade: imagens, discursos e sociabilidade nas praças de João Pessoa - PB

The present work analyses and discusses the sociability in the culture of contemporary cities focusing on the theme of fear and violence in the daily life of squares in the city of João Pessoa-PB. We analyse the relations between places in town and make correlations of history and setting of neighborhoods with the process of urban growth, including recent interventions of public authorities in their public spaces. We observed in this dynamics social attitudes and actions that range from a desired peaceful coexistence to social exclusion. Thus, we discuss, based upon the imaginary vision, that the squares exist as a locus of equality, citizenship and political, and, also, as a space for everyone, despite the existence of symbolic forces working towards segregation and privatization, ruled in fear and violence. We aimed to investigate the different symbolic logics from the issue of fear and violence that allows the exclusion and the inclusion of groups and individuals in the quotidian use of public spaces of contemporary cities. We believe that the social action is demarcated by violence ant its corollary, the social fear, and operates based on the logic of a relational game always confrontational but experienced in different forms according to the social segment or group under study. We used a qualitative and quantitative methodology relating data and statistical analysis with categories created for the understanding of subjective factors. Our analyses combined ethnographic elements, periodics research and images of the city and its spaces, with the contribution of a survey that allowed comparisions of five squares of neighborhood based on the daily life under investigation. Our proposal was to deepen the investigations related to the public space of contemporary cities, expanding the look on João Pessoa and its cultural dynamics with an analysis of discourses, images, the collective imaginary and the social appropriation of the spaces based on fear and violence. The research accomplished in different areas and the analysis of images and speeches published in newspapers reports, books, advertising etc. allowed the approach of differentiated patterns of sociability in the same urban process. The neighborhoods in study are indeed spatially and economically distanced and the process of creation and construction of squares occurred in very different ways in the respective location. We defend the thesis that neighborhood community squares provide reinvigorated spaces and public spheres in the urban process and in the dynamics of sociability in the cities. These squares are also social spaces par excellence for the perception of the logic of individualism and segregation so marked by fear and violence in contemporary cities

Seleção seqüencial de descritores por análise da semântica para recuperação de imagens baseada no conteúdo

With the rapid growth of databases of various types (text, multimedia, etc..), There exist a need to propose methods for ordering, access and retrieve data in a simple and fast way. The images databases, in addition to these needs, require a representation of the images so that the semantic content characteristics are considered. Accordingly, several proposals such as the textual annotations based retrieval has been made. In the annotations approach, the recovery is based on the comparison between the textual description that a user can make of images and descriptions of the images stored in database. Among its drawbacks, it is noted that the textual description is very dependent on the observer, in addition to the computational effort required to describe all the images in database. Another approach is the content based image retrieval - CBIR, where each image is represented by low-level features such as: color, shape, texture, etc. In this sense, the results in the area of CBIR has been very promising. However, the representation of the images semantic by low-level features is an open problem. New algorithms for the extraction of features as well as new methods of indexing have been proposed in the literature. However, these algorithms become increasingly complex. So, doing an analysis, it is natural to ask whether there is a relationship between semantics and low-level features extracted in an image? and if there is a relationship, which descriptors better represent the semantic? which leads us to a new question: how to use descriptors to represent the content of the images?. The work presented in this thesis, proposes a method to analyze the relationship between low-level descriptors and semantics in an attempt to answer the questions before. Still, it was observed that there are three possibilities of indexing images: Using composed characteristic vectors, using parallel and independent index structures (for each descriptor or set of them) and using characteristic vectors sorted in sequential order. Thus, the first two forms have been widely studied and applied in literature, but there were no records of the third way has even been explored. So this thesis also proposes to index using a sequential structure of descriptors and also the order of these descriptors should be based on the relationship that exists between each descriptor and semantics of the users. Finally, the proposed index in this thesis revealed better than the traditional approachs and yet, was showed experimentally that the order in this sequence is important and there is a direct relationship between this order and the relationship of low-level descriptors with the semantics of the users

Compressão auto-adaptativa de imagens coloridas

Image compress consists in represent by small amount of data, without loss a visual quality. Data compression is important when large images are used, for example satellite image. Full color digital images typically use 24 bits to specify the color of each pixel of the Images with 8 bits for each of the primary components, red, green and blue (RGB). Compress an image with three or more bands (multispectral) is fundamental to reduce the transmission time, process time and record time. Because many applications need images, that compression image data is important: medical image, satellite image, sensor etc. In this work a new compression color images method is proposed. This method is based in measure of information of each band. This technique is called by Self-Adaptive Compression (S.A.C.) and each band of image is compressed with a different threshold, for preserve information with better result. SAC do a large compression in large redundancy bands, that is, lower information and soft compression to bands with bigger amount of information. Two image transforms are used in this technique: Discrete Cosine Transform (DCT) and Principal Component Analysis (PCA). Primary step is convert data to new bands without relationship, with PCA. Later Apply DCT in each band. Data Loss is doing when a threshold discarding any coefficients. This threshold is calculated with two elements: PCA result and a parameter user. Parameters user define a compression tax. The system produce three different thresholds, one to each band of image, that is proportional of amount information. For image reconstruction is realized DCT and PCA inverse. SAC was compared with JPEG (Joint Photographic Experts Group) standard and YIQ compression and better results are obtain, in MSE (Mean Square Root). Tests shown that SAC has better quality in hard compressions. With two advantages: (a) like is adaptive is sensible to image type, that is, presents good results to divers images kinds (synthetic, landscapes, people etc., and, (b) it need only one parameters user, that is, just letter human intervention is required

Limiarização contextual automática de imagens coloridas: aplicação na extração de sementes de rodovia

In this paper is a totally automatic strategy proposed to reduce the complexity of patterns ( vegetation, building, soils etc.) that interact with the object 'road' in color images, thus reducing the difficulty of the automatic extraction of this object. The proposed methodology consists of three sequential steps. In the first step the punctual operator is applied for artificiality index computation known as NandA ( Natural and Artificial). The result is an image whose the intensity attribute is the NandA response. The second step consists in automatically thresholding the image obtained in the previous step, resulting in a binary image. This image usually allows the separation between artificial and natural objects. The third step consists in applying a preexisting road seed extraction methodology to the previous generated binary image. Several experiments carried out with real images made the verification of the potential of the proposed methodology possible. The comparison of the obtained result to others obtained by a similar methodology for road seed extraction from gray level images, showed that the main benefit was the drastic reduction of the computational effort.

RR3D: Uma solução para renderização remota de imagens médicas tridimensionais

The visualization of three-dimensional(3D)images is increasigly being sed in the area of medicine, helping physicians diagnose desease. the advances achived in scaners esed for acquisition of these 3d exames, such as computerized tumography(CT) and Magnetic Resonance imaging (MRI), enable the generation of images with higher resolutions, thus, generating files with much larger sizes. Currently, the images of computationally expensive one, and demanding the use of a righ and computer for such task. The direct remote acess of these images thruogh the internet is not efficient also, since all images have to be trasferred to the user´s equipment before the 3D visualization process ca start. with these problems in mind, this work proposes and analyses a solution for the remote redering of 3D medical images, called Remote Rendering (RR3D). In RR3D, the whole hedering process is pefomed a server or a cluster of servers, with high computational power, and only the resulting image is tranferred to the client, still allowing the client to peform operations such as rotations, zoom, etc. the solution was developed using web services written in java and an architecture that uses the scientific visualization packcage paraview, the framework paraviewWeb and the PACS server DCM4CHEE.The solution was tested with two scenarios where the rendering process was performed by a sever with graphics hadwere (GPU) and by a server without GPUs. In the scenarios without GPUs, the soluction was executed in parallel with several number of cores (processing units)dedicated to it. In order to compare our solution to order medical visualization application, a third scenario was esed in the rendering process, was done locally. In all tree scenarios, the solution was tested for different network speeds. The solution solved satisfactorily the problem with the delay in the transfer of the DICOM files, while alowing the use of low and computers as client for visualizing the exams even, tablets and smart phones

Aplicação de programa de análise de imagens na interpretação de fotomicrografias de alta resolução de argilominerais

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Determinação automática da altura de edifícios em imagens aéreas

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Influência da injunção da base na fototriangulação de imagens obtidas com uma unidade de mapeamento móvel

Pós-graduação em Ciências Cartográficas - FCT

A estereoscopia: investigação de processos de aquisição, edição e exibição de imagens estereoscópicas em movimento

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Imagens na literatura: adaptação de Os Maias para a televisão

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Violência e livro didático: um estudo sobre as ilustrações em livros de história

Pós-graduação em Psicologia - FCLAS

Instrumento e método para adquirir sinais e imagens do trato gastrintestinal

This invention relates to an apparatus and method for improving detection by equipment used to detect magnetic susceptibility. The advantages provided are related to different factors, such as reduced instrumentation cost, performance in terms of stability and sensitivity, etc. The devices may be used, for example, to detect magnetic tracers and markers in the gastrointestinal tract of animals and humans. They can be used in research and/or for diagnostic purposes using information related to a variety of parameters of the gastrointestinal tract, such as, pharyngeal and oesophageal transit time, gastric emptying and motility, colonic motility, inter alia.

Avaliação de uma técnica para geração de modelos digitais de superfície utilizando múltiplas imagens

The efficient generation of digital surface model (DSM) from optical images has been explored for many years and the results are dependent on the project characteristics (image resolution, size of overlap between images, among others), of the image matching techniques and the computer capabilities for the image processing. The points generated from image matching have a direct impact on the quality of the DSM and, consequently, influence the need for the costly step of edition. This work aims at assessing experimentally a technique for DSM generation by matching of multiple images (two or more) simultaneously using the vertical line locus method (VLL). The experiments were performed with six images of the urban area of Presidente Prudente/SP, with a ground sample distance (GSD) of approximately 7cm. DSMs of a small area with homogeneous texture, repetitive pattern, moving objects including shadows and trees were generated to assess the quality of the developed procedure. This obtained DSM was compared to cloud points acquired by LASER (Light Amplification by Simulated Emission of Radiation) scanning as wells as with a DSM generated by Leica Photogrammetric Suite (LPS) software. The accomplished results showed that the MDS generated by the implemented technique has a geometric quality compatible with the reference models.