一种全息声元语音编码方法  

Speech Encoding-Holography Sound Cell Encoding

作　　者：李国栋[1] 韩金仓[1] 白顺科 李振东[1] 

机构地区：[1]兰州商业学院信息工程学院,兰州730020 [2]南京步群光电技术有限公司技术部,南京210012

出　　处：《数据采集与处理》2006年第B12期55-58,共4页Journal of Data Acquisition and Processing

基　　金：兰州商业学院高层次人才科研基金(4086)资助项目。

摘　　要：引入了一种新的语音编码方法——全息声元编码,该编码技术将语音节分解为一个声元序列,其中每一个声元均包含原始语音节的主要信息,而全体声元则完全包含原始语音节的全部信息。实验表明,该编码方法具有无浮点运算、运算量小、容错能力强等优点,即使在随机损失75%编码数据的情况下也能很好地重建原始语音信号。A new speech coding method called the holography sound cell encoding is introduced based on audio holography cell transformation. By the new method, speech signal is transformed into a series of holography cells compressed in time axis and similar to each other in waveform. Experimental results show that the method has non-floating-point calculation, small computation and strong fault-tolerant. Even 75% audio holography cells are lost in transmitting link, the original speech signal is reconstructed with a good quality.

关 键 词：语音编码 全息声元 容错 通信 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]