基于嵌入式系统的眼动跟踪技术研究进展

Research Progress on Eye Tracking Technology Based on Embedded Systems

摘要: 眼动跟踪技术作为人机交互、医疗辅助、智能驾驶和虚拟现实等领域的重要支撑手段，近年来在嵌入式系统发展的推动下展现出显著的应用潜力。与传统平台相比，嵌入式系统具有体积小、功耗低和响应快等特点，更加适合部署于智能眼镜、可穿戴设备等对实时性与能效要求严格的场景。本文系统梳理了眼动跟踪的基本原理及人眼运动类型，分析了嵌入式系统的典型组成框架，并列举了当前主流的眼动跟踪方法。针对系统集成中存在的功耗瓶颈、计算资源限制以及数据互操作性差等问题，总结了相关优化策略及代表性成果。结合发展趋势，提出未来研究应关注低功耗算法设计、隐私保护与合规性等方向，以促进眼动跟踪技术从实验研究向大众化、低功耗、高集成的实际应用转变。

Abstract: Eye-tracking technology, as an important support tool in areas such as human-computer interaction, medical assistance, intelligent driving and virtual reality, has demonstrated significant application potential in recent years, driven by the development of embedded systems. Compared with traditional platforms, embedded systems have the characteristics of small size, low power consumption and fast response, making them more suitable for deployment in smart glasses, wearables and other scenarios where real-time performance and energy efficiency are strictly required. This paper systematically appears to comprise the basic principles of eye movement tracking and the types of human eye movement, what the analysis tends to support is a typical composition framework of embedded systems, and lists what appears to represent the current mainstream eye movement tracking methods. What seems to emerge from these findings regarding relevant optimization strategies and representative results were largely summarized for problems such as power consumption bottlenecks, computing resource limitations and poor data interoperability in system integration. What the evidence appears to reveal, given the complexity of these theoretical relationships, is that combined with the development trend, it seems to lend support to what may represent that future research should apparently focus on low-power algorithm design, privacy protection and compliance, in order to promote eye tracking technology from what appears to be experimental research to practical application that are popular, low-power, and highly integrated.