摘要
传统的小波能量特征提取方法没有考虑到每个分解频带上能量的时域分布。针对这一缺点,对传统方法进行了改进,提出了一种基于小波能量矩的局部能量提取方法,用于雷达高分辨距离像的特征提取。小波能量矩既反映了每个频带的能量大小,也反映了能量在时域轴的分布情况。目标识别仿真结果表明,基于小波能量矩的特征提取方法在识别率方面的性能优于传统方法。
The temporal domain distribution of energy on each decomposed band is not considered in the conventional feature extracting method based on wavelete energy. To overcome the defect, the conventional method is improved, and a new method to extract local energy based on wavelet energy- moment is presented in this paper. The new method can be used in feature extraction of high range- resolution radar profiles. The energy-moment feature indicates not only the energy level on each band, but also the energy distribution on the temporal domain axis. Simulation results of target recognition show the feature extraction method based on wavelet energy-moment is better than the conventional method in respect of the recognition rates.
出处
《火力与指挥控制》
CSCD
北大核心
2012年第9期82-84,88,共4页
Fire Control & Command Control
关键词
小波分析
能量
能量矩
特征提取
vavelet analysis ,energy,energy-moment ,feature extraction