摘要
鉴于偏微分方程在图像去噪中的原理和应用,针对传统机械振动信号去噪方法的局限性,提出了一种基于小波变换模改进Perona-Malik模型的强噪声信号滤波算法并用于机械振动信号去噪。首先研究了小波阈值去噪和Perona-Malik非线性各向异性扩散滤波模型之间的相关性,其次用小波变换模替代梯度模构建改进的扩散系数,并推导出了基于小波变换模的改进Perona-Malik模型。实验结果表明,与传统去噪方法和基本Perona-Malik模型相比,改进Perona-Malik模型不仅较好地实现了强噪声背景信号有效去噪,而且同时保留了信号细节特征,改进算法抗噪声干扰能力强,去噪之后信号畸变小,改进算法使信噪比平均提高了约3 dB。
Aiming at traditional mechanical vibration signals de-noising method’s limitation,considering partial differential equations’principle and application in image de-noising,a strong noise signal filtering algorithm based on wavelet transform module and modified Perona-Malik model was proposed.Firstly,the correlation between the wavelet threshold de-noising and Perona-Malik nonlinear anisotropic diffusion filtering model was studied.Secondly,wavelet transform module was used to substitute gradient module and construct an improved diffusion coefficient.The modified Perona-Malik model was derived based on wavelet transform module.The test results showed that compared with the traditional de-noising method and the basic Perona-Malik model,the modified Perona-Malik model can not only realize mechanical vibration signals’effective de-noising under strong noise background,but also keep signals’detail features with little signal distortion;it has a strong anti-noise capacity,the new algorithm makes the average SNR increase by about 3 dB.
作者
毋文峰
陈小虎
WU Wenfeng;CHEN Xiaohu(Department of Management Science and Engineering,Officers College of PAP,Chengdu 610213,China;Department of Equipment Management Engineering,Rocket Force University of Engineering,Xi’an 710025,China)
出处
《振动与冲击》
EI
CSCD
北大核心
2018年第17期277-282,共6页
Journal of Vibration and Shock
基金
四川省科技计划项目(2016JY0222)
