| 音频跳变点的分割熵检测算法 |
| |
| 引用本文： | 宣丽萍. 音频跳变点的分割熵检测算法[J]. 黑龙江科技学院学报, 2008, 18(3) |
| |
| 作者姓名： | 宣丽萍 |
| |
| 作者单位： | 黑龙江科技学院,电气与信息工程学院,哈尔滨,150027 |
| |
| 摘 要： | 为了克服基于贝叶斯信息准则(BIC)的音频跳变点检测方法存在的缺陷,提出了音频跳变点的分割熵检测算法.该方法通过检测音频信号内部信息熵的变化趋势来检测跳变点的声学特征.结果表明:该方法可避免因设定固定门限而导致的漏检和数据积累带来的累积误差,能够准确地切分各种声学特性的音频数据,是一种更为直观和有效的跳变点检测方法,具有很高的灵敏度和鲁棒性.
|
| 关 键 词： | 跳变点检测 音频分割 信息熵 |
Acoustic change point detection method based on entropy of segmentation |
| |
| XUAN Liping. Acoustic change point detection method based on entropy of segmentation[J]. Journal of Heilongjiang Institute of Science and Technology, 2008, 18(3) |
| |
| Authors: | XUAN Liping |
| |
| Affiliation: | XUAN Liping(College of Electrical , Information Engineering,Heilongjiang Institute of Science , Technology,Harbin 150027,China) |
| |
| Abstract: | Overcoming the shortcomings arising from use of BIC for acoustic change point detection forces the development of a novel method based on entropy of segmentation for acoustic change point detection.This method allows the detection of acoustic change point by detecting entropy changing trend inside broadcasting data.The results show that this method helps to avoid error accumulation due to establishing fixed threshold.The method is a more intuitionistic and effective method for acoustic change point detectio... |
| |
| Keywords: | acoustic change point detection broadcasting data segmentation entropy |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
