A wavelet theory about online wavelets denoising based on Moving Window and Principal Component Analysis (PCA)

Abstract

In this paper, we have described a general wavelet theory about online wavelet denoising based on Moving Window and Principal Component Analysis (PCA). Using the online lifting scheme of signals and wavelet thresholding in a moving window of dyadic length, we can remove unpleasant or noise errors in the data. Insufficiency of traditional Wavelet denoising in real-time signal processing is discussed. Requirements of online denoising are studied, and a moving window is introduced into traditional Wavelet transform. Genuine images are frequently corrupted by noise from various sources. It has been confirmed to have a better edge-preserving quality than linear filters in certain applications. By using the moving window, an online Wavelet denoising method is recommended. Many different developments are described by the signal extensively used in denoising domain. The simulation results show the success of these improvements for fault diagnosis.