A new wavelet de-nosing algorithm based on reversible transform

Traditional wavelet shrinkage de-noising methods always produce Pseudo-Gibbs oscillations when sensor signals contain jump discontinuity points. In order to solve this problem, this paper presents a wavelet shrinkage de-noising algorithm based on reversible translation. The algorithm can eliminate Pseudo-Gibbs oscillations by removing the discontinuity points. Its rationality is proven theoretically and the most appropriate mother wavelet is Harr wavelet. At the same time, the simulation data and real course signal show that this algorithm can effectively eliminate Pseudo-Gibbs oscillations and improve denoising SNR.