On the Usage of Context for Augmented Reality Visualization

Augmented reality is a user interface paradigm, which mixes virtual and physical information in a single output. Augmenting virtual information on top of physical information at interactive framerates causes a number of issues. These include poor perception of distances, image pollution and attention direction. In order to address these problems, this dissertation explores the usage of information that is not the focus of the current task but provides a situational reference. This information is often referred to as contextual information or context. This dissertation addresses multiple uses of context for augmented environments. It attempts to encourage the usage of context in augmented reality. It does so by presenting a series of techniques that address a variety of problems, ranging from scenegraph styling to human visual attention. All of the presented techniques are successful in attaining their goal. The work presented in this dissertation allows the styling of scenegraphs on the fly based on tagged context. It also improves the perception of depth in augmented environments for single occlusion setups. Finally, by carefully manipulating context it is capable of directing the human visual attention to desired portions in the image.