Wenfeng Wu / Quanzhou University of Information Engineering
Xiaohu Chen / Quanzhou University of Information Engineering
A new nonlinear diffusion filtering algorithm in wavelet domain was put forward for mechanical vibration signal denoising by integrating the wavelet threshold denoising algorithm and the nonlinear diffusion filtering algorithm. At first, the partial differential equation theories were introduced and the Perona-Malik model, as one of classical nonlinear diffusion filtering models (algorithms), was brought in. From the image denoising applications, it can be analyzed and inferred that both the wavelet threshold denoising algorithm and the Perona-Malik model have their respective disadvantages. However there were a few interrelations and complementarities between the wavelet threshold denoising algorithm and the Perona-Malik model. Thus by integrating these two algorithms, a new nonlinear diffusion filtering algorithm in wavelet domain was put forward for signal denoising. Furthermore its algorithm principle was elaborated with a flow chart. By comparison with the solo wavelet threshold denoising algorithm and the solo Perona-Malik model, the new nonlinear diffusion filtering algorithm in wavelet domain can not only realize signal denoising but also reserve signal features without signal distortions. The simulation experiments indicate that this new algorithm is a perfect synthesis of wavelet transform and Perona-Malik model with better denoising performance and better anti-noise robustness.