CUICA - uma ferramenta para suportar a convergência da TV digital com a WEB através da criação de conteúdo digital

AUTOR(ES)
DATA DE PUBLICAÇÃO

2007

RESUMO

A collaborative virtual environment can be dened as a single virtual reality space hared by multiple participants connected from diferent hosts. However, most collaborative existing systems restrict the communication among the participants to text messages or audio communication. The natural means of human communication are richer than this. Other efects of coordinated visual and touch feedback play also a fundamental role and create a more realistic experience to the users. Motivated by the possibility of navigating and interacting remotely in a collaborative virtual environment, as well as simulating and sharing the perception of contact forces in diferent scenarios, an architecture for collaborative virtual environments for training, using force feedback devices, was designed, implemented and tested. The proposed architecture is composed of 4 components: the Device Manager, that makes possible the force feedback devices mapping; the Virtual Reality Environment, where the 3D application resides; the Collaboration Manager, that consists of a collaboration protocol for data sharing and controlling during the training sessions; and the Device Controller, responsible for the low-level routines needed to interact with the haptic device. The manager-type components, in particular improve the architecture _exibility and were implemented in such a way to be able to support extensions. To analyze the system performance, a significative number of experimental evaluations was conducted. The performance evaluations do not restrict themselves to the ideal environment conditions provided by a local area network, but also were applied to diferent network setups, where diferent bandwidth, delay and packet loss values were simulated. Based on these experimental data, a similarity metric was dened to relate the processing time spent by the rendering process of the hosts (the masters that leads the training, and the trainee_s that follows the activities executed by the master) during the training exercise. The similarity values, obtained experimentally, were then compared to those obtained by a subjective analysis realized by the users. The results validated the architecture and demonstrated that the proposed methodology was useful to elaborate a detailed performance analysis of the system. Even in face of a wide variation in the processing times, obtained in nanoseconds, and consequently di_culty in the analysis of the packet loss impact on results, it was possible to successfully evaluate the efects of bandwidth and delay in the application performance, making use of similarity metrics that varied according to a range of di_erent percentual values of tolerance. The definition of a rigid metric for the users subjective analysis was also suitable, particularly for comparing the experimental results to the users perceptions, in attempt to _nd out a possible range of trusting values for the tolerance parameters. KEYWORDS: Performance Analysis . Collaborative Virtual Environment Force Feedback . Training

ASSUNTO(S)

televisÃo digital - dissertaÇÕes internet - dissertaÇÕes sistemas de informacao

ACESSO AO ARTIGO

Documentos Relacionados