Voluntary blink control is vital in eye-controlled human–computer interaction (ECHCI). Pressure sensors, known for their low power consumption and strong resistance to interference, are suitable for blink detection. Existing research has predominantly focused on sensor materials and signal processing, and studies on the characteristics and optimized design of voluntary blink actions in ECHCI remain relatively limited. Therefore, this study presents a blink detection system based on pressure sensors and investigates the efficiency of voluntary blink actions. We examined the effects of two independent variables, blink count (single, double, and triple) and blink side (bilateral and unilateral), on recognition accuracy and three time-related variables. Subsequently, the subjective workload and evaluation experiments were conducted for each blink action. The results indicate that the blink count has a more pronounced impact on interaction efficiency and comfort, with continuous blink actions exceeding three blinks not being recommended. Bilateral blinks were perceived as more natural and comfortable than unilateral blinks. Moreover, we discovered that the participants exhibited greater attentiveness and focus when performing double bilateral blink actions, making it suitable for important commands. Based on the experimental results, the top three recommended voluntary blink actions are single bilateral blink (SB), double bilateral blink (DB), and single unilateral blink (SU). Finally, a test involving controlling the movement of a toy car using the recommended blink actions validates the usability of the actions and detection system. These findings provide valuable guidance for the design and optimization of voluntary-blink interaction systems, further enhancing this interaction method’s feasibility.

中文翻译：

---

基于压力传感器的自愿眨眼交互的设计建议

自愿眨眼控制在眼控人机交互（ECHCI）中至关重要。压力传感器以功耗低、抗干扰能力强而著称，适合眨眼检测。现有的研究主要集中在传感器材料和信号处理方面，对于 ECHCI 中自主眨眼动作的特征和优化设计的研究仍然相对有限。因此，本研究提出了一种基于压力传感器的眨眼检测系统，并研究了自愿眨眼动作的效率。我们研究了两个自变量：眨眼次数（单次、两次和三次）和眨眼次数（双边和单边）对识别准确性和三个时间相关变量的影响。随后，对每个眨眼动作进行主观负荷和评价实验。结果表明，眨眼次数对交互效率和舒适度有更明显的影响，不建议连续眨眼动作超过三次。双侧眨眼被认为比单侧眨眼更自然、更舒适。此外，我们发现参与者在执行双边眨眼动作时表现出更高的注意力和注意力，使其适合执行重要命令。根据实验结果，推荐的前三种自愿眨眼动作是单侧眨眼（SB）、双侧眨眼（DB）和单侧眨眼（SU）。最后，涉及使用推荐的眨眼动作控制玩具车运动的测试验证了动作和检测系统的可用性。这些发现为自愿眨眼交互系统的设计和优化提供了有价值的指导，进一步增强了这种交互方法的可行性。