Gai Guanghong,Qu Liangsheng,Li Chongsheng
(School of Mechanical Engineering,Xi'an Jiaotong University,Xi'an 710049,China)
Abstract: A novel 1D signals denoising method based on the second generation wavelet was presented in digital signal processors(DSP) chip. The signals are only decomposed into one multiresolution level，and then the detail coefficients and the approximation coefficients are modified via soft thresholding and inversely synthesized. The advantages of the method are lower computational complexity and reduced memory requirement, so it is suitable to the real-time implementation using DSP chip. Moreover, in order to reduce the execution time, the program structure was optimized by parallel operations. The implementation was verified by the simulation signals. Applied to measured signals from laboratory it can reduce the white noise from signals effectively. Results obtained show that this method is a feasible DSP solution.
Keywords: second generation wavelet;signal denoising;digital signal processors; real-time implementation