摘要
有效地从含有噪声的非平稳信号中提取特征是进行非平稳信号分类等研究的基础。应用流域算法,对含有高斯白噪声的非平稳信号的时频分布图进行分割,并根据能量占优的准则对其合并,提出了一种基于能量的特征提取方法。仿真结果表明该方法能有效地提取特征量,且对高斯白噪声具有很好的抗噪性能。
The important basis on the classification and target recognization of nonstationary signal is how to effectively extract feature from it in the gaussian white noise ambience. The watershed algorithm is used to segment the time-frequency distribution of a signal and combine segmentations according energy maximum criterion, then a feature extraction method based on energy is proposed. Simulative examples demonstrate that it can effectively extract feature and has a good anti-noise performance to gaussian white noise.
出处
《红外与激光工程》
EI
CSCD
北大核心
2004年第3期296-299,共4页
Infrared and Laser Engineering
关键词
流域算法
时频分析
能量
Algorithms
Energy dissipation
Signal processing
White noise
