Will it Move?: Indoor Scene Characterization for Hologram Stability in Mobile AR
Will it Move?: Indoor Scene Characterization for Hologram Stability in Mobile AR
复制标题
DOI：
10.1145/3446382.3449026
复制
发表时间：
2021-02
期刊：
影响因子：
--
通讯作者：
T. Scargill;Shreya Hurli;Jiasi Chen;M. Gorlatova
中科院分区：
文献类型：
--
作者：
T. Scargill;Shreya Hurli;Jiasi Chen;M. Gorlatova
移动增强现实（AR）通过使虚拟内容（全息图）与现实世界的看法相结合来提供沉浸式的体验。当用户放置全息图时，通常可以预期像真实对象一样，它仍然位于同一位置。但是，位置误差经常是由于环境映射和设备定位不正确而发生的，这在很大程度上取决于现场自然视觉特征的性质。在此演示中，我们介绍了SceneIt，这是基于全息图位置误差幅度的预测移动AR的第一个视觉环境评级系统。 SceneIt允许用户确定将虚拟内容放置在其环境中是否会显着地漂移，而无需他们放置该内容。它表明，对于给定的视觉环境来说，位置误差的严重性是可以预测的，并且可以以足够高的精度和低潜伏期来计算该预测，以便在移动AR应用程序中有用。
Mobile Augmented Reality (AR) provides immersive experiences by aligning virtual content (holograms) with a view of the real world. When a user places a hologram it is usually expected that like a real object, it remains in the same place. However, positional errors frequently occur due to inaccurate environment mapping and device localization, to a large extent determined by the properties of natural visual features in the scene. In this demonstration we present SceneIt, the first visual environment rating system for mobile AR based on predictions of hologram positional error magnitude. SceneIt allows users to determine if virtual content placed in their environment will drift noticeably out of position, without requiring them to place that content. It shows that the severity of positional error for a given visual environment is predictable, and that this prediction can be calculated with sufficiently high accuracy and low latency to be useful in mobile AR applications.