On Signal Denoising Using Quasi Real-Time Wavelet Algorithm In Dsp Embedded System

Jun Zhao,Weiguo Liu
DOI: https://doi.org/10.1109/ICOSP.2008.4697065
2008-01-01
Abstract:In this paper, a quasi real-time Wavelet denoising algorithm was used to improve the online denoising effect based on wavelet transform, which is realized on the DSP embedded platform. The simulation and analysis, compared with off-line wavelet algorithm, were discussed. And the plaform was designed to fulfil the requirement of quasi real-time wavelet algorithm. With the quasi real-time wavelet algorithm, the platform reduced the noise effectively form sigma(original) = 1.00000 to sigma(denoised) = 0.65589. The minimum processing period of the platform is 0.25ms with 256 points moving data window. The effect was obvious in application of the signal denoising, which is used to measure liquid rate successfully.
What problem does this paper attempt to address?