基于时频特征的Wi-Fi手势识别技术  

Wi-Fi gesture recognition technology based on time-frequency features

作　　者：任梦恬 田增山[1] 蒋青[1] REN Mengtian;TIAN Zengshan;JIANG Qing(School of Communication and Information Engineering,Chongqing University of Posts and Telecommunications,Chongqing 400065,P.R.China)

机构地区：[1]重庆邮电大学通信与信息工程学院,重庆400065

出　　处：《重庆邮电大学学报（自然科学版）》2021年第3期458-465,共8页Journal of Chongqing University of Posts and Telecommunications(Natural Science Edition)

基　　金：国家自然科学基金(61771083,61704015);长江学者和创新团队发展计划(IRT1299);重庆市研究生科研创新项目(CYS17221)。

摘　　要：随着人工智能的快速发展,手势识别已经成为重点关注和研究的对象,利用Wi-Fi信号进行手势识别的技术无需额外的设备以及光照的条件,逐渐成为手势识别的主流。针对传统基于时域统计特征的方法,将经过噪声抑制后的信号进行短时傅里叶变换(short-time Fourier transform,STFT)构建手势信号的频域特征,结合时域和频域特性选择动作持续时间、频谱熵等特征作为特征值,采用支持向量机(support vector machine,SVM)的分类方法对特征值进行训练并完成手势识别。实验结果表明,在复杂的室内环境与室外空旷环境均能有效对手势进行识别,手势的识别率达90%以上。With the rapid development of artificial intelligence,gesture recognition has become the focus of attention and research.Gesture recognition using Wi-Fi signals has become the mainstream of gesture recognition because it does not require additional equipment and lighting conditions.For gesture recognition method based on traditional time domain statistical features,in this paper,the short-time Fourier transform(STFT)of the denoised signal is used to construct the frequency domain features.Then,combining time domain and frequency domain features,selecting features such as action duration and spectral entropy as features,and the features are performed by using the support vector machine(SVM)classification method to train and complete gesture recognition.The experimental results show that this paper can effectively identify the gestures in the complex indoor environment and outdoor environment.The recognition rate of gestures is over 90%.

关 键 词：时频域特征值 短时傅里叶变换 WI-FI 手势识别 

分 类 号：TP393.1[自动化与计算机技术—计算机应用技术]