Implementation of wavelet shrinkage de-noising based on DSP
Chen Xiaoli, Tian Mao · 2004
Wavelet shrinkage has been a universal wavelet de-noising method in many signal-processing applications. It always consists of three steps: a forward wavelet transform, a shrinkage de-noising and an inverse wavelet transform. This paper addresses wavelet shrinkage de-noising algorithm based on DSP. The paper firstly introduces the principle of discrete wavelet transform and soft thresholding. Then it describes the implementation platform of wavelet shrinkage de-noising - analog devices DSP (digital signal processor) ADSP21060, which is a high performance signal processor. And the paper proposes to realize the wavelet shrinkage de-noising algorithm based on DSP instead of PC in order to be more suitable for real time system. Moreover, we advocate the realization method of thresholding algorithm based on DSP. In addition, the experiments show that wavelet shrinkage de-noising based on DSP is satisfying.