语谱分析的FPGA实现
doi: 10.3724/SP.J.1146.2010.01005
Realization of Spectrogram Analysis with FPGA
-
摘要: 语谱是语音信号短时时频分析结果的图形显示，能够清楚地揭示语音信号的时变频谱特性，反映语音信号的动态频谱特性和时域变化特性，在语音信号的分析以及语音学的研究中具有重要的价值。该文利用短时傅里叶变换(Short Time Fourier Transform，STFT)方法对语音信号进行分析，采用了一种按时域抽取基-4 FFT算法实现流水线结构的蝶形运算单元，在 FPGA上实现了语音信号的语谱分析，并利用非线性映射算法实现了语谱的VGA显示。
-
关键词:
- 语谱 /
- 时频分析 /
- 短时傅里叶变换(STFT) /
- 蝶形运算单元 /
- FPGA
Abstract: Spectrogram is a kind of graphic representation of time-frequency analysis of speech signals. Because spectrogram can be adopted to demonstrate the dynamic characteristics of time varying speech signals both in frequency domain and time domain, it has been widely employed in speech analysis and phonetic researches. This paper realizes a spectrogram on Field Programmable Gate Array (FPGA) with Short Time Fourier Transform (STFT) method by using a line structure butterfly process unit which adopted Radix4 decimation in time FFT algorithm. The spectrogram is also displayed on VGA displayer by adopting a non-linear mapping algorithm. -
杨海钢, 孙嘉斌, 王慰. FPGA器件设计技术发展综述[J].电子与信息学报.2010, 32(3):714-727浏览[5]Yang Hai-gang, Sun Jia-bin, and Wang Wei. An overview to FPGA device design technologies[J].Journal of Electronics Information Technology.2010, 32(3):714-727[6]朱银兵, 许江宁, 崔国恒, 马建国. 罗兰C周期识别时频分析方法研究[J].电子与信息学报.2010, 32(3):732-736浏览[7]Zhu Yin-bing, Xu Jiang-ning, Cui Guo-heng, and Ma Jian-guo. Time-frequency methods for period identification of Loran-C[J].Journal of Electronics Information Technology.2010, 32(3):732-736 -
计量
- 文章访问数: 3560
- HTML全文浏览量: 132
- PDF下载量: 1229
- 被引次数: 0
下载:
