【摘要】 皮电信号(GSR)是心理测试的重要参数指标。信号的奇异点包含着皮电信号的幅度变化、突变时间及持续时间等重要特征,且信号具有非平稳,超低频等特点。传统的信号处理方法不能有效地对其进行滤波并进行奇异点检测,甚至将包含重要信息的奇异点信号滤除。采用小波变换处理皮电信号,选用Daubechies小波系将皮电信号分解为一个近似皮电信号和相应的细节成分,完成对皮电信号的去噪处理,且对几种小波消噪方式进行了比较,得出适合于皮电信号的消噪方法。最大限度地保留了奇异点的完整性。消噪信号利用李普希兹(Lipschitz)指数,通过模量极大值和微分对相对应心理变化引起的皮电信号第一类间断点和第二类间断点进行了检测。通过间断点推断皮电信号的突变时间和持续时间,实现心理特征自动检测用于心理评估。更多还原

【Abstract】 Galvanic skin response signal(GSR) is one of the most important index of psychological test,the singularity of GSR contains many important characteristics such as varies in amplitude,start time and duration of singularity.The conventional signal processing method can’t filter the noise and detect the singularity effectively, even the singularity were filtered because of the non-stationary and ultra low frequency.Wavelet transform was selected to process GSR.GSR was decomposed into an approximation components and corresponding detail components using wavelet of Daubechies series and the effect is obvious.The suitable de-noising way of GSR is selected by comparing several ways.Completeness of singularity is retained in maximum.The first type break and the second type break of de-noised signal are detected by calculating the maxima and differential using Lipschtiz.Start time and duration time of GSR are inferred,eigenvalues are used for psychological assessment after obtained by auto -detection.更多还原