本地滞后对协同增强现实系统性能影响研究

 一致性控制是协同增强现实研究核心问题，本地滞后算法可有效消除协同过程中的短期不一致，但如何选择一个合适的本地滞后值是该算法的关键问题。文章构建了一个具体的协同增强现实系统，并在不同本地滞后值下对该系统性能进行研究。实验结果表明本地滞后值（0~600 ms）对系统性能的影响是非线性的，这些影响可分为三段：0~100，100~500，500 ms以上，当本地滞后值小于100 ms时，本地滞后对系统性能没有影响，当本地滞后值从100 ms增加到500 ms时，本地滞后对系统的性能影响随滞后值变大而变大，但被试均能完成任务，当本地滞后值超过500 ms时，系统性能急剧变坏，一些被试不能顺利完成任务。

Effects of Local-Lag Mechanism on Task Performance in a Collaborative Augmented Reality System

Consistency maintenance is a kernel problem to Collaborative Augmented Reality (CAR) research. In this area, local lag mechanism has been proposed to eliminate short term inconsistency, while how to select a proper lag value is a key problem to the mechanism. In this paper, a particular CAR system is presented, and an experiment was conducted to investigate the effects of lag value on the task performance of the system. Experimental results indicate that the effect of lag value on task performance is not linear, and it could be separated into three segments by two dividing points: 100 ms and 500 ms. When the lag is less than 100 ms, the effect of lag is negligible. When the lag increases from 100 ms to 500 ms, the negative effect of lag increases with the lag, but all participants can accomplish the task. When the lag exceeds 500 ms, the task performance decreases rapidly, and some participants even cannot accomplish the task.