Non-photorealistic 3D video-avatar
Daniel M. Tokunaga, Ricardo Nakamura, Romero Tori · 2009
The digital games industry is always looking for innovation in user experience. A new trend in this field is the use of Augmented Reality (AR) techniques for intuitive, novel game interfaces [Bernardes Jr et al. 2008]. Among the several technologies related to AR, video avatars are one of the most attractive for games, because they allow the insertion of the game player's image in the game. Furthermore, the availability of commodity stereo cameras makes it feasible to employ 3D video avatars in consumer games. However, when a non-photorealistic rendering style is required for design reasons (often the case for games), the use of a conventional video avatar, even with 3D information, results in visual inconsistencies. This work presents a new approach to enable nonphotorealistic rendering of a real-time 3D video avatar. The project builds upon a 3D video avatar system designed for teleconferencing over Internet 2 for educational purposes, extending it with this new rendering method. The expected result is a reduction in the visual and cognitive mismatch between player image and synthetic environment, allowing for a more immersive experience.