Soft strain sensors pose great potential for emerging human–machine interfaces. However, their real-world applications have been limited due to challenges such as low reproducibility, susceptibility to environmental noise, and short lifetimes, which are attributed to nanotechnologies, including microfabrication techniques. In this study, we present a computer vision-based optical strain (CVOS) sensor system that integrates computer vision with streamlined microfabrication techniques to overcome these challenges and facilitate real-time multiaxial strain mapping. The proposed CVOS sensor consists of an easily fabricated soft silicone substrate with micro-markers and a tiny camera for highly sensitive marker detection. Real-time multiaxial strain mapping allows for measuring and distinguishing complex multi-directional strain patterns, providing the proposed CVOS sensor with higher scalability. Our results indicate that the proposed CVOS sensor is a promising approach for the development of highly sensitive and versatile human–machine interfaces that can operate long-term under real-world conditions.

软应变传感器为新兴的人机界面带来了巨大的潜力。然而，由于纳米技术（包括微加工技术）带来的低重复性、易受环境噪声影响和寿命短等挑战，它们的实际应用受到限制。在这项研究中，我们提出了一种基于计算机视觉的光学应变（CVOS）传感器系统，该系统将计算机视觉与简化的微加工技术相结合，以克服这些挑战并促进实时多轴应变映射。所提出的 CVOS 传感器由易于制造的带有微型标记的软硅胶基板和用于高灵敏度标记检测的微型相机组成。实时多轴应变映射可以测量和区分复杂的多方向应变模式，为所提出的 CVOS 传感器提供更高的可扩展性。我们的结果表明，所提出的 CVOS 传感器是开发高灵敏度和多功能人机界面的一种有前途的方法，该界面可以在现实条件下长期运行。