Optimal Placement of Virtual Content in Physical Space

Research output: Book/anthology/dissertation/reportPh.D. thesisResearch

In recent years, a lot of research has been done to move user interfaces away from the conventional desktop metaphor, and incorporate the physical environment as a means of input and output. Tracking technology is evolving to allow increasingly accurate explicit input, e.g., through gestures, but also implicit input, e.g., by tracking users' viewpoints. However, in contrast to desktop interfaces, room-scale user interfaces are in the same realm as the physical environment. Physical spaces are not necessarily built with the user interface in mind, but have an inherent purpose. Examples include office spaces, living rooms and manufacturing environments. Physical and virtual entities might now compete for the user's attention. Advances in automated projector calibration methods enable virtual content to be co-located with and bound to physical entities, to merge it seamlessly with physical environments. However, comparatively little work has been done on incorporating the usage of the physical space in the calibration process.

This dissertation explores placement of virtual content in room-scale user interfaces. Besides being bound to physical surfaces, the content placement has a strong focus on user-related criteria like visibility. The contribution of this work consists of several novel content placement approaches in room-scale physical environments. The core of the contribution is to integrate user viewing behavior in the optimization process for content placement. Three systems with different levels of automation for content placement are presented:
MeetAlive enables manual placement of projected virtual content anywhere in a meeting room. A room-scale WIMP interface allows users to move content and thereby ensure its visibility. HeatSpace enables installers of physical displays to find optimal positions for displays in a physical environment. The system measures user viewing behavior and the physical environment to semi-automatically propose optimal display placements as to ensure visibility to users. OptiSpace enables fully automatic placement of 3D projection mapping content. With this, developers of room-scale user interfaces can define spatial constraints for virtual 3D content and deploy applications in unknown target environments. In addition to the three content placement systems, this dissertation presents Velt, a framework that facilitates the development of research prototypes, which are based on multi RGBD camera networks.
Original languageEnglish
Number of pages176
StateSubmitted - 30 Jun 2018

See relations at Aarhus University Citationformats

ID: 128531570