| 改进AR模型特征提取及分类 |
| |
| 引用本文： | 尚小晶,吴忠伟,徐成波.改进AR模型特征提取及分类[J].吉林建筑工程学院学报,2014(3):74-77. |
| |
| 作者姓名： | 尚小晶 吴忠伟 徐成波 |
| |
| 作者单位： | 吉林建筑大学城建学院,长春130111 |
| |
| 基金项目： | 吉林省教育厅“十二五”科学技术研究项目（2014596）. |
| |
| 摘 要： | 目前,表面肌电信号（sEMG）是手势动作识别研究的重要信号源.本文以肌电信号为对象,从非平稳与非线性的角度出发,采用ICA独立成分分析和经验模式分解的方法,消除表面肌电信号中的工频干扰,对处理后的信号建立AR模型.将模型系数作为信号的特征,对6种手势动作进行模式识别.实验表明,该方法获得的特征具有较好的分类效果.
|
| 关 键 词： | 表面肌电信号 独立成分分析 经验模式分解 |
The Improved AR Model Feature Extraction and Classification |
| |
| Affiliation: | SHANG Xiao - jing, WU Zhong - wei, XU Cheng - bo ( The City College of Jilin Jianzhu University, Changchun, China 130111) |
| |
| Abstract: | At present the surface myoelectric signal （sEMG） is an important source of gesture recognition. The article uses EMG signal is non - stationary and nonlinear object from the point of view, by using the method of decom- position of ICA independent component analysis and empirical models to eliminate power interference of EMG, establishes the AR model to the processed signal. The model coefficients are used as the signal features, the pattern recognition of 6 kinds of gestures. Experiments show that, the method obtained has better classification effect. |
| |
| Keywords: | surface EMG signal independent component analysis empirical mode decomposition |
| 本文献已被 维普 等数据库收录！ |
|
