文章摘要
刘春花,付强,杨家玮,颜永红.音频信号截幅失真的检测与修复[J].声学技术,2009,(4):498~502
音频信号截幅失真的检测与修复
Detection and restoration of clipping distortion in audio signal
投稿时间:2008-07-22  修订日期:2008-10-17
DOI:
中文关键词: 数字截幅  模拟截幅  AR模型  LSAR
英文关键词: digital clipping  analogue clipping  AR modeling  LSAR
基金项目:国家高技术研究发展计划(2006AA010102,2006AA01Z195);国家重点基础研究发展规划项目计划(2004CB318106);国家自然科学基金(10574140,60535030)资助
作者单位E-mail
刘春花 西安电子科技大学通信工程学院, 西安, 710071 liuchunhua2008@gmail.com 
付强 中国科学院卢学研究所, 北京, 100190  
杨家玮 西安电子科技大学通信工程学院, 西安, 710071  
颜永红 中国科学院卢学研究所, 北京, 100190  
摘要点击次数: 1290
全文下载次数: 2665
中文摘要:
      音频信号截幅失真的修复是信号修复工作的重要组成部分,目的是使得信号恢复完备性,以便于对其进行进一步研究。提出了一种音频信号截幅失真的检测与修复方法。首先通过研究截幅失真特性,描述了基于音频信号峰值大小而进行的检测算法,并创造性地将基于AR模型以及最小均方误差的LSAR算法应用于修复算法中。随后给出应用该算法后得到的实验结果,以及在截幅失真测试集上得到的信噪比及PESQ值改善的统计,由上述实验数据可以验证,该方法对不同采样率截幅失真的修复均有一定效果,达到了音频修复的要求,提高了音频质量。
英文摘要:
      Clipping distortion in audio signal is an important part of signal restoration, whose aim is to make the audio integrative in order to do further research on it. This paper presents a method for detection and restoration of clipping distortions in audio signals. It begins with a description of the method for the detection of clipping distortions based on the audio’s amplitude peak value, and creatively uses a least square AR (LSAR) repairing method based on the AR modeling and the theory of the minimum mean square error for the restoration of the clipping distortions. Subsequently, the results obtained by this method are shown, and the statistical results acquired by the method effecting on the clipping distortions test collection-the signal to noise ratio (SNR) and the value of the Perceptual evaluation of speech quality (PESQ) are presented. Obviously, the method is efficient when it processes on the clipping distortions with different sample rates, achieves the demands of audio restoration, and improves the audio quality.
查看全文   查看/发表评论  下载PDF阅读器
关闭