首页 | 本学科首页   官方微博 | 高级检索  
     检索      

参数音频编码算法的改进
引用本文:陶峻,陈健.参数音频编码算法的改进[J].通信技术,2002(12):40-43.
作者姓名:陶峻  陈健
作者单位:上海交通大学DSP技术中心,上海,200030
摘    要:参数音频编码是一种基于产生音频信号的参数模型的有效编码方法。最近已用于很低码率的音频编码中,如MPEG-4Version2音频标准HILN。这里提出HILN参数音频编码的改进算法。改进算法采用了自适应的基于倒谱的基频检测算法,利用倒谱和频谱特性,保证了大范围内基频提取的效率和准确性。在正弦波合成中采用相位匹配合成算法,保证了帧间参数的连续性。改进算法还对残差信号进行预处理,去除不必要的干扰分量,最大限度重建原音频信号,提高了合成音频信号的质量。

关 键 词:正弦模型  噪声模型  HILN  参数音频编码
修稿时间:2002年7月9日

Improved Parametric Audio Coding Algorithm
Tao Jun Chen Jian.Improved Parametric Audio Coding Algorithm[J].Communications Technology,2002(12):40-43.
Authors:Tao Jun Chen Jian
Abstract:Parametric Audio Coding is based on the decomposition of an audio signal into components, which are described by appropriate source models a nd represented by model parameters. It is utilized in very low bit rate audio co ding. In this paper, a modified algorithm for HILN, which was proposed in 2000 b y MPEG-4 Version2 is presented. An adaptive fundamental frequency estimation ba sed on cepstrum is proposed, which can extract the fundamental frequency and its harmonics within a large frequency domain efficiently. To keep the continuity b etween different frames, phase matching algorithm is also utilized. And a pre-p rocessing is performed on the residual signal to clean off the residual sinusoid s and improved the audio quality.
Keywords:sinusoid model  residual model  HILN  parametric audio codin g
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号