一种改进的基音周期提取算法  

Improved Algorithm for Pitch Period Detection

作　　者：赵祎 张盛[1] 林孝康[1] 

机构地区：[1]清华大学深圳研究生院深圳市信息科学与技术重点实验室,深圳518055

出　　处：《数据采集与处理》2014年第2期304-308,共5页Journal of Data Acquisition and Processing

摘　　要：基音周期的提取在语音信号处理领域有着广泛的应用。受基于归一化自相关函数基音周期提取算法和多带激励(Multi-band excitation,MBE)声码器中基音检测算法的启发,本文提出了一种改进的基音周期提取算法。该算法主要由预处理、时域基音粗估、基音平滑、时变滤波搜索、小数基音周期估计等5个部分组成。实验表明:该算法能达到更高的搜索准确度,得到更加平滑的基音周期曲线;与传统自相关检测算法相比,该算法有很好的抗噪性。The extraction of pitch period has a wide range of applications in the field of speech signal processing. Inspired by traditional autocorrelation algorithm and the pitch detection method that used in multi-band excitation （MBE） vocoder, an improved algorithm is put for- ward for pitch period extraction. This algorithm has five parts： pre-process, pitch rough esti- mation in time domain, pitch smoothing, search with time variable filter, and decimal pitch es- timation. Experimental results show that this new algorithm can achieve higher accuracy and compared with traditional autocorrelation algorithm, this approach has a better noise immunity

关 键 词：基音周期 归一化自相关 基音平滑 时变滤波 

分 类 号：TN912.3[电子电信—通信与信息系统]