摘要
本文提出了基于小波分析技术的弹头表面纹理特征提取方法 ,对于传统小波分析的边界失真进行了讨论 ,提出了一种无损重构技术 ,该方法相对于以前所用的采取数据两端补零和两端对称延长的方法能够更好的消除边界失真。将该种方法用于弹头发射痕迹的纹理特征提取中 。
In this paper, a method of extracting the texture feature of warhead based on wavelet analysis is put forwarded.The errors at the boundary of the traditional wavelet analysis are discussed.Based on the polynomial fit, the method to eliminate the errors is advanced.Compared with the method of adding zero at both ends of the data and extending both ends symmetrically,it can eliminate the errors all the better.This method, applied in extracting the texture feature of the marks on the warhead,has won the satisfactory result.
出处
《仪器仪表学报》
EI
CAS
CSCD
北大核心
2001年第z2期202-203,共2页
Chinese Journal of Scientific Instrument
关键词
纹理特征提取
小波分析
弹头发射痕迹
Texture feature extraction Wavelet analysis Fired warhead marks