Augmented reality (AR) is the process of overlaying additional information (usually visual) onto our perception of the world in real time. In many cases, the goal is to have visual augmentations blend with the real world in a visually coherent manner. This visual coherence may provide the user with stronger visual cues on the location, shape and characteristics of the virtual objects, and of the interactions between them and the real objects. Furthermore, visual coherence may greatly contribute to achieving the main goals of certain specific AR experiences, such as engaging the users in an interactive, immersive 3D narrative or videogame, or enhancing telepresence in an AR CSCW environment.
Currently, most AR applications use graphical technologies that do not blend well with the view of the physical world. In many AR experiences, virtual objects have illumination and shading that are too simple and/or incoherent with the illumination of nearby real objects. Also, visual artifacts common to the blending of virtual and real objects (such as aliasing or disparity between color gamuts) create further gaps in realism.
Our goal is to improve the state of the art by emphasizing the use of real-time global illumination (GI) techniques in AR experiences. This entails finding ways of obtaining illumination information from the real environment, then using that information as input for photorealistic rendering techniques, and finally blending the real and virtual content in a coherent manner.