摘要
该文在Bark子波的构造的基础上 ,提出一种改进的子波变换。该子波变换在Bark子波变换的非线性映射中引入伸缩尺度参数 ,以及改变Bark子波母函数的相关参数 ,实现了更为灵活的频域划分。然后基于改进的子波变换进行水声目标特征提取 ,及分类实验。实验表明 。
This paper describes a new wavelet transform and its application in underwater acoustic targets classification. The new wavelet transform is the improvement of Bark wavelet transform. It introduces a scale parameter into the nonlinear mopping and changed relevant coefficient involved in Bark wavelet transform. It can analyse signal in frequency domain freely. Then feature extraction is performed based on the new wavelet transform. The classification experiments show that the feature based on the new wavelet transform is superior to that based on Bark wavelet transform.
出处
《计算机仿真》
CSCD
2005年第2期26-28,50,共4页
Computer Simulation
关键词
子波变换
特征提取
目标分类
Wavelet transform
Feature extraction
Target classification
