基于静音识别的改进型ADDPCM语音压缩算法的研究  

Study on the speech compression algorithm of the ADDPCM based on voice active detection

作　　者：王涌[1] 贾立新[1] 何剑春[1] 

机构地区：[1]浙江工业大学信息工程学院,浙江杭州310032

出　　处：《浙江工业大学学报》2006年第6期660-663,共4页Journal of Zhejiang University of Technology

摘　　要：针对语音信号振幅较大时传统压缩算法的跟踪性较差的情况,提出了一种新型的改进型自适应ADDPCM语音压缩算法.同时,根据语音信号不连续的特点,压缩算法具有静音识别功能,进一步提高了语音信号的压缩率.压缩算法在8 kHz的A/D采样速率下,将实时传输和处理的语音数据减少到了3 kb/s的数据流量,并且解码后的语音恢复信号具有良好的可懂度和较高的语音能量,取得了良好的语音压缩效果.在语音信号压缩算法研究的基础上,利用片上系统单片机C8051F005实现了语音信号的压缩编码和解码,并进行了效果的比较.实验表明,该语音压缩算法有效降低了编码率,提高了传输带宽的利用率,并且恢复的语音信号在振幅较大时取得较好的动态跟踪效果.The tracking ability of traditional speech compression algorithm is weak when the swing of voice signal is high. A new-style and improved ADDPCM is proposed in this paper. According to the characteristic of discontinuous voice signal, this compression algorithm provides the function of voice active detection and improves the compress ratio of voice signal. With the sampling of 8 kHz, the speech compression algorithm reduces the data stream to 3 kb/s. The decoded voice signal is good at understandability and has high voice energy. The compressing and decoding processes of voice signal are carried out under SOC（System on chip）based on the speech compression algorithm. The experimentation has indicated that the improved ADDPCM voice signal compression algorithm has higher compresses ratio and is good at dynamic tracking when the swing of voice signal is high.

关 键 词：语音压缩算法 静音识别 片上系统单片机 

分 类 号：TP391.42[自动化与计算机技术—计算机应用技术]