Powered by OpenAIRE graph
Found an issue? Give us feedback
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/ Universidade de Lisb...arrow_drop_down
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
UTL Repository
Master thesis . 2019
Data sources: UTL Repository
versions View all 2 versions
addClaim

Redes neuronais no âmbito da realidade virtual.

Authors: Correia, Paulo Jorge Duarte;

Redes neuronais no âmbito da realidade virtual.

Abstract

O objectivo deste trabalho é propor um modelo para a obtenção da estrutura tridimensional de uma cena baseado na utilização de redes neuronais que possa ser aplicado no desenvolvimento de sistemas de Telepresença. O problema que se coloca é a obtenção da estrutura tridimensional de uma cena a partir da análise das suas projecções numa sequência de imagens bidimensionais. Como tal , trata-se de um problema de Visão por Computador, que envolve três fases: detecção de características presentes nas imagens, estabelecimento de correspondências entre os pontos característicos e obtenção da estrutura tridimensional. Neste trabalho são estudadas e analisadas Tecnologias de Informação consideradas inovadoras; a Realidade Virtual e as Redes Neuronais, No caso da Realidade Virtual, são descritas as suas características e potencialidades e é feito um levantamento de alguns campos de aplicação. De seguida, é efectuado um paralelismo com as modernas Tecnologias de Informação e Sistemas de Informação e são especificados alguns dos efeitos que poderá exercer sobre as Unidades Económicas. É efectuado um levantamento de alguma pesquisa levada a cabo no domínio da Visão por Computador como forma de introdução a esta problemática. De seguida, são apresentadas as redes neuronais para a resolução do problema. Para o estabelecimento de correspondências é utilizada uma rede de Hopfield, Define-se uma função de custo que incorpora as restrições do problema, a qual irá ser minimizada pela rede. Cada neurónio representa uma possível correspondência entre um ponto de uma imagem e um ponto da imagem seguinte. Por fim, apresenta-se a rede neuronal para a obtenção da estrutura tridimensional a partir das correspondências. É seguido o princípio da máxima rigidez, à semelhança do que acontece com o sistema visual humano. Existe um modelo interno da estrutura tridimensional, que será actualizado gradualmente, permitindo a existência de alguns desvios de rigidez.

The purpose of this work is to propose a model for the recovery of the tridimensional structure of a scene based on neural networks. This model may be used to develop Telepresence sysíems. The problem that vve face is to determine the 3-D structure of a scene from an analysis of its projections on a sequence of 2-D images. This is a Computer Vision problem that involves three phases: features detection, matching of the feature points, and computation of the 3-D structure. In this work we study and analyse Information Technologies which are considered innovative: Virtual Reality and Neural Networks. In the case of Virtual Reality, we describe some of its characteristics and potentials and we present some of its possible applications. Next, we compare Virtual Reality to the modern Information Technologies and Information Systems and we suggest some of the possible effects of its use by Economic Units. We refer some of the work done in the fíeld of Computer Vision so that we can show some of the problems that researchers in this field have to face. Next, we state the neural networks that will be used to solve the problem. For the matching phase we suggest the use of an Hopfield neural network. We state a cost function that will represent the constraints on the solution, which will be minimised by the network. Each neurone represents a possible match between a point of an image and a point in the next image. Then, we present the neural network that will be used to compute the 3-D structure from the correspondences established. That computation is constrained by the maximal rigidity principie, a process similar to the way we see. There is an internai model of the 3-D structure that will be updated at each instant, allowing for some deviations from rigidity.

Instituto Superior de Economia e Gestão

N/A

Country
Portugal
Keywords

Estrutura tridimensional, Visão por computador, Tecnologias de informação, Structure, Redes neuronais, Computer vision, Realidade virtual, Information technologies, Neural networks, Virtual reality

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    0
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Average
    OpenAIRE UsageCounts
    Usage byUsageCounts
    visibility views 7
    download downloads 7
  • 7
    views
    7
    downloads
    Powered byOpenAIRE UsageCounts
Powered by OpenAIRE graph
Found an issue? Give us feedback
visibility
download
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
views
OpenAIRE UsageCountsViews provided by UsageCounts
downloads
OpenAIRE UsageCountsDownloads provided by UsageCounts
0
Average
Average
Average
7
7
Green
Related to Research communities