基于短时傅立叶变换的脉象信号的模式识别方法

针对脉象信号的非平稳特性,采用全极点滑动窗递归算法,对15例吸毒者和15例正常人脉象信号的离散短时功率谱进行了分析。在得到每一例脉搏波的短时功率谱后,应用奇异值分解有效地提取特征矢量,然后进行模糊c-均值聚类,受测者全部予以正确识别。研究结果表明,基于短时傅立叶变换的奇异值分解方法是一种稳定、有效的特征提取方法;同样,运用模糊c-均值聚类算法不需要模式的先验知识,分类结果正确率较高,简便实用。

The Pulse Signals Pattern Identification Method Based on Short-time Fourier Transform

To fully utilize the nonstationary character of pulse signal,an efficient recursive algorithm with all-pole moving-windows is used to analyze the discrete short-time power spectra of pulse signals for 15 heroin addicts and 15 healthy persons.After obtaining a short-time power spectra of every pulse wave,singular value decomposition is then used to extract feature vector for pattern identification,then the fuzzy c-means cluster algorithm is also used.All subjects characterized by the feature vectors are identified correctly.The research result shows that it is a stable and efficient method for extracting features based on singular value decomposition of short-time Fourier transform.Also using the fuzzy c-mean cluster algorithm does not need the apriori knowledge of pattern identification,and accuracy of the classified result is high and FCM is simply practical.