| 背景噪声下的端点检测算法的研究 |
| |
| 引用本文： | 周娜,赵振东,张莹.背景噪声下的端点检测算法的研究[J].通信技术,2008,41(7). |
| |
| 作者姓名： | 周娜 赵振东 张莹 |
| |
| 作者单位： | 1. 华北电力大学电子与通信工程系,河北保定,071003
2. 东北财经大学津桥商学院,辽宁大连,116000 |
| |
| 摘 要： | 噪声环境下的端点检测在语音信号分析和识别中占有重要地位.文中将分形理论中的分形记盒维数应用到端点检测算法中,采用了基于分形记盒维数与短时能零比相结合的端点检测算法,以分形记盒维数为主要判决条件,并在判决门限的设定上采用了自适应机制.仿真结果表明,基于分形记盒维数和短时能零比相结合的端点检测算法在白噪声和粉红色噪声下具有很好的准确率和鲁棒性.
|
| 关 键 词： | 端点检测 能零比 分形记盒维数 自适应机制 背景噪声 端点检测算法 研究 Background Noise Based 鲁棒性 准确率 色噪声 白噪声 仿真结果 适应机制 判决门限 条件 结合 短时能零比 应用 记盒维数 分形理论 地位 识别 |
Endpoint Detection Methods Based on Background Noise |
| |
| ZHOU Na,ZHAO Zhen-dong,ZHANG Ying.Endpoint Detection Methods Based on Background Noise[J].Communications Technology,2008,41(7). |
| |
| Authors: | ZHOU Na ZHAO Zhen-dong ZHANG Ying |
| |
| Affiliation: | ZHOU Na ,ZHAO Zhen-dong ,ZHANG Ying (Dept. of Electronic & Communication Engineering,North China Electric Power University,Baoding Hebei 071003,China,Kingbridge Business College,Dongbei university of Finance & Economics,Dalian Liaoning 116600,China) |
| |
| Abstract: | Endpoint detection in noisy environment plays an important role in speech analysis and recognition. In this paper, the box-counting dimension in the fractal theory is used in the endpoint detection. The classical threshold method is improved, and the endpoint detection algorithm based on the box-counting dimensions and short-time energy-zero-ratio is employed. The box-counting dimension is taken as the main decision condition, and the self-adaptive mechanism is used to set the decision threshold. The simula... |
| |
| Keywords: | endpoint detection energy-zero-ratio box-counting dimensions self-adaptive mechanism |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
