Visualization and recognition of context information in virtual and augmented reality

Abstract

In this thesis, the visualization and recognition system of the context information in immersive reality is proposed. Three simulation-based empirical experiments demonstrate the effectiveness of the visualization of the context information in safety, procedural learning, and visual engagement. In environmental safety, context information is described as the vehicle's 3D position in the street situation, and out-of-view visualization with the context information is built on the unified framework of recognition and visualization. In procedural learning, the 3D position and the virtual devices' manual are the context information to transfer the user for the training contents. The place to focus on at every moment is defined as the context information in the visual engagement. The empirical experiment shows that immersion and the context information's situation-awareness have inversely proportional to environmental safety. In procedural learning, two visualization methods, the annotation, and the tutor, were compared for transferring the knowledge of the training contents, and annotation shows better in the two tasks among three. In the visual engagement experiment, discipline shows better results in terms of gaze distribution but not in terms of usability. The results can be used in the field in terms of informing designers of visualization of the context information on what types of visual instruction are best for different task purposes.