Augmented reality technology is attracting increased attention, given the popularity of the smartphone. The utilisation of the smartphone with AR technology is also enabling mobile augmented reality (MAR) to become accessible. Stereoscopic vision offers users the benefit of depth perception, which can help improve user memory. Therefore, this study aims to investigate the effect on knowledge retention using MAR with stereoscopic vision. An experiment to compare a stereoscopic-based MAR application and a paper-based manual was designed to test the participants’ knowledge retention on a product part and disassembly process. The developed MAR adopted a smartphone and headset case to provide a stereoscopic view to the user. The experiment consisted of both pre-test and post-test phases where the pre-test phase examined the participant’s knowledge about the product, and the post-test phase focused on how much information the participants could recall. The post-test phase was conducted after 48 h following the pre-test phase since long-term memory required several hours in which to stabilise. The results showed that the MAR group had an advantage over the paper-based manual group for both information retentions having a mean score of 8.89 from the MAR group, compared with 6.33 for the paper-based manual group. Moreover, the MAR group was observed to have fewer errors in the post-test with a mean score of 0.53, whereas the paper-based achieved a score of 2.2. This result indicated that the MAR group and the paper-based manual group had equivalent performance in the completion time without the assistance tool. In addition, MAR had a better mean score for the subjective feedback (usefulness = 4.47, ease of use 4.18 and satisfaction = 4.07).

鉴于智能手机的普及，增强现实技术正在引起越来越多的关注。带有AR技术的智能手机的使用还使移动增强现实（MAR）变得可访问。立体视觉为用户提供了深度感知的好处，可以帮助改善用户的记忆力。因此，本研究旨在探讨使用立体视觉的MAR对知识保留的影响。设计了一个比较基于立体视觉的MAR应用程序和基于纸质手册的实验，以测试参与者对产品零件和拆卸过程的知识保留。研发的MAR采用了智能手机和耳机保护套，为用户提供了立体感。实验包括测试前和测试后两个阶段，其中测试前阶段检查参与者对产品的知识，而测试后阶段则关注参与者可以回忆多少信息。由于长期记忆需要数小时的时间才能稳定下来，因此在预测试阶段之后的48小时后进行后测试阶段。结果表明，MAR组比基于纸质手册组的优势在于，MAR组的平均信息得分为8.89，而基于纸质手册组的信息保留率为6.33。此外，MAR组在测试后的错误更少，平均得分为0.53，而纸质得分为2.2。该结果表明，MAR组和纸质手册组在没有辅助工具的情况下，在完成时间上具有同等的表现。此外，MAR对主观反馈的平均评分更高（有用度= 4.47，易用性4.18，满意度= 4.07）。