📄 中文摘要
研究提出了一种新方法，通过提取肌肉信号中的有用信息来应对噪声干扰。肌肉传感器在控制设备和跟踪运动中发挥重要作用，但随机静态信号常常掩盖真实信号，导致失败。新特征MMNF通过分析信号能量的分布，有效忽略了大量随机噪声。在强静态信号干扰的测试中，MMNF的错误率约为5-10%，而许多常见特征的错误率超过20%。此外，将MMNF与基本的直方图测量和简单的幅度计数相结合，进一步提高了识别的准确性。
一种新颖的特征提取方法用于稳健的肌电图模式识别
出处: A Novel Feature Extraction for Robust EMG Pattern Recognition
发布: 2026年2月12日
📄 English Summary
A Novel Feature Extraction for Robust EMG Pattern Recognition
A novel method has been proposed to extract useful information from muscle signals to combat noise interference. Muscle sensors are crucial for controlling devices and tracking movement, but random static often obscures the true signal, leading to failures. The new feature, MMNF, analyzes the distribution of signal energy, effectively ignoring substantial random noise. In tests with strong static interference, MMNF achieved an error rate of approximately 5-10%, while many common features exceeded 20%. Furthermore, combining MMNF with a basic histogram measure and simple amplitude counting further enhanced recognition accuracy.