New Methods for Maintaining Good Stereoscopic Viewing During Interaction with Large-Scale Head-Tracked Displays

虚拟现实（VR）旨在通过计算机生成的视觉和声音将用户感知到人工世界。 因此，VR提供了一种新的人机交互模式，可以提高用户对设计、数据可视化、培训、教育和医学等不同应用领域中的3D信息空间的理解和交互。 立体VR系统通过向每只眼睛呈现单独的透视图像来生成图像。 结果，用户感知到看起来存在于物理显示表面前面和后面的单个真实3D图像。 虽然立体显示是VR系统中的一个重要和常见的功能，但需要进一步的研究，因为当用户在虚拟环境中穿行和操纵时，立体观看引起的关注超出了单视场VR系统中引起的关注。 这对于扩展的虚拟环境尤其如此，其中场景包含以覆盖几个数量级的尺度呈现的几何细节。 在这样的环境中，用户需要放大和缩小以在详细视图和全局视图之间移动。 本项目将研究在这种条件下保持良好的立体视觉条件的技术。 PI将专注于立体头跟踪显示系统，如虚拟工作台（区别于显示表面安装在用户头部的头戴式显示器）。 该项目的预期影响将是更好地理解和实施基于对假眼分离和头部跟踪失真的彻底几何和分析的头部跟踪显示器;用于自动观察光学调节和自动观察位置调节的新的和改进的技术，以保持良好的立体观察条件;系统研究了上述方法与VR应用相关几何属性的兼容性。

Virtual reality (VR) aims to perceptually place the user in an artificial world through computer-generated sight and sound. VR thus offers a new human-computer interaction paradigm which can improve the user's understanding of and interaction with 3D information spaces in diverse applications domains such as design, data visualization, training, education, and medicine. Stereoscopic VR systems generate imagery by presenting a separate perspective image to each eye. As a result, the user perceives a single, true 3D image that appears to exist in front of and behind the physical display surface. While stereoscopic display is an important and common feature in VR systems, further research is needed because stereoscopic viewing raises concerns beyond those raised in monoscopic VR systems as the user travels through and manipulates the virtual environment. This is especially true for extended virtual environments where the scene contains rendered geometric detail at scales covering several orders of magnitude. In such environments, users need to zoom in and out to move between detailed and global views. This project will investigate techniques for maintaining good stereoscopic viewing conditions under such conditions. The PIs will focus on stereoscopic Head-Tracked Display systems such as the virtual workbench (as distinguished from head-mounted displays in which the display surface is mounted on the user's head). The expected impact of this project will be better understanding and implementations of head-tracked displays based on a thorough geometric and analytic analysis of false eye separation and head-tracking distortions; new and improved techniques for automatic view optics adjustments and automatic view position adjustments to maintain good stereoscopic viewing conditions; and a systematic study of the compatibility between the above methods and the relevant geometric attributes of VR applications.