Uma Estratégia para uso de múltiplas câmeras em sistemas de Realidade Aumentada baseados em marcadores fiduciais

• Ano de defesa: 2011
• Autor(a) principal: Arantes, Webert Vieira
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Uberlândia; BR; Programa de Pós-graduação em Engenharia Elétrica; Engenharias; UFU
• Palavras-chave em Português: Interface não convencional; Augmented reality; Interface unconventional; CNPQ::ENGENHARIAS::ENGENHARIA ELETRICA
• Link de acesso: https://repositorio.ufu.br/handle/123456789/14460; https://doi.org/10.14393/ufu.di.2011.50

If compared with other areas of research Augmented Reality shows are fairly recent, and as such subject to new scientific research and technological innovations. According to Kirner and Siscoutto [2007] Augmented Reality environments makes easier the action of the user in three-dimensional space by increasing the system performance and allowing for better interaction techniques. In these environments it is common to use only one camera to capture the real scene and to allow interaction with the system. Therefore, in applications where it is presented a large number of virtual information, at the time of interaction with these data is a great risk of instability in the scene caused by the occlusion of markers, and confusion in data visualization. In addition, the area of interaction with virtual objects is limited to around the user, making the action on the virtual objects located further away. This study proposes the use of more than one camera on the development of these systems and thereby causes an increased interaction area with virtual objects in the scene, making easier the use of markers and enabling a better "cleaning" of environment displayed. The user has the option of presenting the fiducial markers not only the main camera, but the other cameras available. The verification will occur which marker was used by analyzing the camera image which was presented but the corresponding actions will be performed on objects in the scene presented the main camera, allowing the interaction to occur without interference in the environment. In implementing the prototype was used GoblinXNA platform, an open-source platform written in C # on the Microsoft XNA Game Studio and the library ALVAR. To support this context, was used a questionnaire to the participants of the experiments. Through analysis of these could infer that the proposed strategy is feasible and enables an expansion of the area of interaction with virtual objects in augmented reality systems.