Miami University.
Swiss Federal Institute of Technology.
IEEE Trans Vis Comput Graph. 2014 Apr;20(4):579-87. doi: 10.1109/TVCG.2014.34.
Redirected walking algorithms imperceptibly rotate a virtual scene about users of immersive virtual environment systems in order to guide them away from tracking area boundaries. Ideally, these distortions permit users to explore large unbounded virtual worlds while walking naturally within a physically limited space. Many potential virtual worlds are composed of corridors, passageways, or aisles. Assuming users are not expected to walk through walls or other objects within the virtual world, these constrained worlds limit the directions of travel and as well as the number of opportunities to change direction. The resulting differences in user movement characteristics within the physical world have an impact on redirected walking algorithm performance. This work presents a comparison of generalized RDW algorithm performance within a constrained virtual world. In contrast to previous studies involving unconstrained virtual worlds, experimental results indicate that the steer-to-orbit keeps users in a smaller area than the steer-to-center algorithm. Moreover, in comparison to steer-to-center, steer-to-orbit is shown to reduce potential wall contacts by over 29%.
转向步行算法可以将沉浸式虚拟现实系统用户周围的虚拟场景进行不可察觉的旋转，从而引导他们远离跟踪区域边界。理想情况下，这些扭曲可以使使用者在物理空间有限的情况下自然行走，从而探索大型无限的虚拟世界。许多潜在的虚拟世界都是由走廊、通道或过道组成的。假设使用者不期望在虚拟世界中穿过墙壁或其他物体，那么这些受限制的世界限制了行进的方向，以及改变方向的机会数量。因此，物理世界中使用者运动特征的差异会对转向步行算法的性能产生影响。本工作比较了受限虚拟世界中广义 RDW 算法的性能。与之前涉及无约束虚拟世界的研究相比，实验结果表明，与转向中心算法相比，转向轨道算法可以使用户保持在较小的区域内。此外，与转向中心算法相比，转向轨道算法被证明可以减少超过 29%的潜在墙壁接触。
