基于Matlab的基音周期检测  

Pitch Cycle Detection Based on Matlab

作　　者：吴和生 WU Hesheng(Jiangsu Higher Vocational School of Judicial Police,Zhenjiang 212003,China)

机构地区：[1]江苏省司法警官高等职业学校,江苏镇江212003

出　　处：《电声技术》2023年第12期34-36,共3页Audio Engineering

基　　金：江苏省监狱协会一般课题(JSJYXH22J13)。

摘　　要：基音周期检测是语音信号处理的重要内容,在语音合成、编码和识别中发挥着重要作用。短时自相关函数法和倒谱法是基音周期检测的常用算法。为对比短时自相关函数法和倒谱法的检测性能,在分析两种算法基本原理的基础上,采用Matlab仿真平台对两种算法建模。仿真结果表明,倒谱法与短时自相关函数法估算出的基音周期精度相近,均有较好的基音周期检测效果。Pitch cycle detection is an important aspect of speech signal processing,playing a crucial role in speech synthesis,encoding,and recognition.Short time autocorrelation function method and cepstrum method are commonly used algorithms for gene cycle detection.To compare the detection performance of the short-time autocorrelation function method and the cepstrum method,based on the analysis of the basic principles of the two algorithms,Matlab simulation platform was used to model the two algorithms.The simulation results showed that the accuracy of the pitch period estimated by the cepstrum method and the short-time autocorrelation function method was similar,and both had good pitch period detection effects.

关 键 词：基音周期检测 短时自相关函数法 倒谱法 MATLAB 

分 类 号：TN912.2[电子电信—通信与信息系统]