3D Multi-user interactive visualization with a shared large-scale display

Abstract

When the multiple users interact with a virtual environment on a largescale display there are several issues that need to be addressed to facilitate the interaction. In the thesis, three main topics for collaborative visualization are discussed; display setup, interactive visualization, and visual fatigue. The problems that the author is trying to address in this thesis are how multiple users can interact with a shared large-scale display depending on the display setups and how they can interact with the shared visualization in a way that doesn’t lead to visual fatigue.

The first user study (Chapter 3) explores the display setups for multi-user interaction with a shared large-display. The author describes the design of the three main display setups (a shared view, a split screen, and a split screen with navigation information) and a demonstration using these setups. The user study found that the split screen and the split screen with navigation information can improve users’ confidence and reduce frustration level and are more preferred than a shared view. However, a shared view can still provide effective interaction and collaboration and the display setups cannot have a large impact on usability and workload.

From the first study, the author employed a shared view for multi-user interactive visualization with a shared large-scale display due to the advantages of the shared view. To improve interactive visualization with a shared view for multiple users, the author designed and conducted the second user study (Chapter 4). A conventional interaction technique, the mean tracking method, was not effective for more than three users. In order to overcome the limitation of the current multi-user interactive visualization techniques, two interactive visualization techniques (the Object Shift Technique and Activity-based Weighted Mean Tracking method) were developed and were evaluated in the second user study. The Object Shift Technique translates the virtual objects in the opposite direction of movement of the Point of View (PoV) and the Activity-based Weighted Mean Tracking method assigns the higher weight to active users in comparison with stationary users to determine the location of the PoV. The results of the user study showed that these techniques can support collaboration, improve interactivity, and provide similar visual discomfort compared to the conventional method.

The third study (Chapter 5) describes how to reduce visual fatigue for 3D stereoscopic visualization with a single point of view (PoV). When multiple users interact with 3D stereoscopic VR using multi-user interactive visualization techniques and they are close to the virtual objects, they can perceive 3D visual fatigue from the large disparity. To reduce the 3D visual fatigue, an Adaptive Interpupillary Distance (Adaptive IPD) adjustment technique was developed. To evaluate the Adaptive IPD method, the author compared to traditional 3D stereoscopic and the monoscopic visualization techniques. Through the user experiments, the author was able to confirm that the proposed method can reduce visual discomfort, yet maintain compelling depth perception as the result provided the most preferable 3D stereoscopic visualization experience.

For these studies, the author developed a software framework and designed a set of experiments (Chapter 6). The framework architecture that contains the three main ideas are described. A demonstration application for multidimensional decision making was developed using the framework.

The primary contributions of this thesis include a literature review of multiuser interaction with a shared large-scale display, deeper insights into three display setups for multi-user interaction, development of the Object Shift Techniques, the Activity-based Weighted Mean Tracking method, and the Adaptive Interpupillary Distance Adjustment technique, the evaluation of the three novel interaction techniques, development of a framework for supporting a multi-user interaction with a shared large-scale display and its application to multi-dimensional decision making VR system.