Authoring and presenting situated media in augmented and virtual reality.

摘要

This dissertation investigates the research challenges in authoring and presenting situated media in augmented and virtual reality. We use the term situated media to refer to multimedia and hypermedia that are embedded in the surrounding spatial environment. Linking information to physical objects and locations provides the ability to access and interact with information in context, using head-worn and hand-held displays. It potentially eases interaction with the physical world, enhances perception, and helps with users' tasks.; We have developed authoring techniques that enable non-programmer users to create a variety of augmented and virtual reality experiences, ranging from stand-alone multimedia to richly linked hypermedia narratives embedded in the users' surroundings, which we call situated hypermedia. Depending on the task at hand, and the level of precision required, users can use stationary desktop-based or mobile tablet-based tools for authoring, to take advantage of the different input devices and interaction techniques available on these platforms. We collaborated with content creators who used our desktop-based authoring system and related presentation tools at different stages of their development to author situated hypermedia for course projects, and present informal feedback from these experiences.; An essential complement to authoring is the ability to present the created content in an interactive and intuitive way to the users. To accomplish this, we have developed presentation techniques that make it possible for users to interact with situated media and navigate in large situated media spaces. Our augmenting pictures technique and animated timelines register and superimpose historic images onto real or virtual objects and provide a means of interacting with them, to make it possible for users to gain a better understanding of historic events and places. Our lifting, tilting, shifting and scaling techniques enable users to navigate and visualize complex hypermedia structures distributed across large distances. Using our techniques, users can selectively "pick up" interesting parts of a complex situated hypermedia structure, lift or tilt them to view otherwise occluded nodes and links, as well as shift and scale the structure for closer examination. We also present the results of a formal user study that evaluates several of these techniques.