HCC: Small: Depth Perception in Near- and Medium-Field Augmented Reality

项目摘要HCC：小号：增强现实（AR）是一种技术，用户可以看到计算机生成的虚拟对象叠加在他们的世界视图上。 在制造、物流、维护、态势感知、抬头地图、培训、旅游、考古、戏剧、舞蹈、建筑和图像引导手术等领域，已经提出并开发了大量引人注目的AR应用。 所有这些AR应用的共同挑战是将虚拟对象放置在真实的世界中的特定深度。 由于固有的工程限制，AR显示器不显示具有与真实的对象相同的深度线索的虚拟对象。 此外，一些AR应用还涉及X射线视觉，即显示不透明表面后面的虚拟物体。 对于这些应用，深度线索必然冲突，但仍然产生有用的信息。 这导致了该项目正在研究的科学问题：（1）深度感知如何与AR的冲突深度线索一起运作，以及（2）需要开发和验证可以更有效地传达AR深度的新方法。 调查人员正在通过两个主要项目轨道解决这些问题。 第一个轨道是经验性地研究如何计算机生成的图像的深度被认为是近场距离内的手臂？的范围和中场距离为1.5至30米。 第二个轨道是实现和经验测试几个新的基于眼动跟踪器的AR深度呈现方法。 这些包括聚散度深度渲染，其中眼睛的聚散度角度控制哪些AR对象可见，以及模拟景深，其中与聚散度角度不同深度的对象被模糊。 这模拟了具有固定焦深的AR显示器中的景深模糊。 项目？的实验采用经验的方法，从历史悠久的深度知觉研究和使用控制条件，允许验证通过比较与此文献。

Project AbstractHCC: Small: Depth Perception in Near- and Medium-Field Augmented RealityAugmented reality (AR) is a technology where users see computer-generated, virtual objects superimposed on their view of the world. A large number of compelling AR applications have been proposed and developed in areas such as manufacturing, logistics, maintenance, situation awareness, heads-up maps, training, tourism, archeology, theater, choreography, architecture, and image-guided surgery. A common challenge for all of these AR applications is placing virtual objects at a specific depth in the real world. Because of inherent engineering limitations, AR displays do not show virtual objects with the same depth cues as real objects. Furthermore, some AR applications involve x-ray vision, which is the display of virtual objects that exist behind opaque surfaces. For these applications, the depth cues necessarily conflict, but still yield useful information. This leads to the scientific problems that this project is studying: (1) how depth perception operates with the conflicting depth cues of AR, and (2) new methods that can more effectively convey AR depth need to be developed and validated. The investigators are addressing these problems through two major project tracks. The first track is empirically studying how the depth of computer-generated images is perceived for near-field distances within arm?s reach and medium-field distances of 1.5 to 30 meters. The second track is implementing and empirically testing several new eye tracker-based AR depth presentation methods. These include vergence depth rendering, where the vergence angle of the eyes controls which AR objects are visible, and simulated depth-of-field, where objects at different depths than the vergence angle are blurred. This simulates depth-of-field blurring in an AR display with a fixed focal depth. The project?s experiments adopt empirical methods from the long history of depth perception research and use control conditions that allow validation through comparison with this literature.