混合码激励／声激励低码率语音编码算法  

A COMBINED CODE-EXCITED AND VOICE-EXCITED LOW BIT RATE SPEECH CODING SCHEME

作　　者：韦岗[1] 欧阳景正[1] 

机构地区：[1]华南理工大学无线电与自动控制研究所

出　　处：《华南理工大学学报（自然科学版）》1994年第3期24-31,共8页Journal of South China University of Technology(Natural Science Edition)

基　　金：广东省自然科学基金

摘　　要：本文提出了一种混合低码率编码算法。原始语音首先经低通及高通滤波器分解成基带及高带信号，然后对基带信号采用高质量的码激励线性预测编码，对高带信号则采用码率低的改进声激励线性预测编码，以达到高质量低码率。实验结果表明，本算法在码率为４．０ｋｂ／ｓ时，合成语音的质量与传统的４．８ｋｂ／ｓ码激励线性预测编码相当，而计算量只有其１／５，是一种在低码率下可达到高质量的编码方案。A combined code-excited and voice-excited speech coding scheme is proposed in this paper. In the algorithm, the speech signal is firstly divided into two parts in frequency domain, of which the base-band signal is coded with code-excited linear prediction and the high-band signal is coded with voice-excited linear prediction. The proposed algorithm has lower comPlexity than the original code-excited linear prediction, and can produce satisfactory synthesized speech.

关 键 词：信源编码 语音信号处理 言语传递 

分 类 号：TN912.3[电子电信—通信与信息系统]