摘要
针对传统脉搏周期检测方法的准确率低、误差大的问题,提出一种基于平均幅度差函数AMDF(Average Magnitude Difference Function)的脉搏周期检测改进算法。根据干扰谷点的分布范围提出脉搏信号的噪声阈值,并运用新提出的阈值削波函数排除噪声干扰,获得了更精确的脉搏周期检测结果。实验结果表明,该算法运算复杂度低,能准确地提取脉搏周期,错判率低,稳健性好。
To solve the problem of low accuracy and great error of the traditional pulse period detection method, the paper puts forward an AMDF-based （ average magnitude difference function） pulse period detection improved algorithm. According to the distribution of the interference valley points, it proposes the threshold of pulse signal, by applying the newly proposed threshold clipping function to exclude noise dis- turbance, hence obtaining more accurate pulse period detection result. Experiment results indicate that the complexity of the computing of the algorithm is low and it can precisely extract pulse period, is low at error rate and good at steadiness.
出处
《计算机应用与软件》
CSCD
北大核心
2014年第6期52-54,共3页
Computer Applications and Software
基金
江西省自然科学基金项目(20114BAB211012)
江西省教育厅科技项目(GJJ12275)
关键词
脉搏信号
特征点提取
周期检测
平均幅度差函数
自相关函数
Pulse signal Feature point extraction Period detection AMDF（ average magnitude difference function） ACF（ auto corre-
