基于Hilbert-Huang变换基音周期检测算法  

Hilbert-Huang Transform Based Pitch Detection

作　　者：闫磊[1] 李永红[1] 杜力力[1] 

机构地区：[1]中北大学信息与通信工程学院,山西太原030051

出　　处：《科技信息》2013年第7期203-204,共2页Science & Technology Information

摘　　要：介绍了一种与传统算法相比有更好的自适应性和优越性的基音周期检测算法,Hilbert-Huang变换不需要对语音信号进行短时平稳假设,检测精度高,适应范围广,帧长大大增加;与小波变换相比,Hilbert-Huang变换依据信号本身的信息对信号进行分解,随信号本身变化而变化,表现了信号内含的真实物理信息,具有更好的自适应性和优越性。A comparison with the traditional algorithm have a better self-adaptalbility anti superiority of pitch detection algorithm. Hilbert- Huang Transfiorm to the voice signal does not require a smooth short-term assumptions, detection of high precision, to adapt to a wide range of significant increase in frame length; compared with the wavelet transform, Hilbert-Huang transforn their information signal based on the signal decomposition, with the signal itself changes, the performance of a signal containing information on the real physical aml has better self-adaptability and superiority.

关 键 词：HILBERT-HUANG变换 基音周期检测 瞬时能量 

分 类 号：TU311.3[建筑科学—结构工程]