Research On Wavelet Denoising Algorithm And Wavelet Hardware Realization

Wavelet denoising is a very popular and front topic in the field of signal processing. The interference of noise can make a great impact on the quality of signals and images, which will result in a big problem of the following signal processing such as feature extraction, signal detection and speech recognition etc. Noise will cause an error or even a mistake for the processing result.In recent years, the wavelet theory has obtained the extremely rapid development. The wavelet analysis is internationally recognized signal acquisition and processing of information in the high-tech field and it is the multi-disciplinary attention hot spot and signal processing front topic. For the characteristics of its low entropy, multi-analysis, relativity removal and flexible bases, the wavelet transform has become a powerful tool in the field of signal denoising.With the wavelet threshold denoising theory proposed by D.L.Donoho, a new threshold function is presented in this paper, which is based on the hypothesis that the noise is additional and gaussian. The new threshold function has two order continuous derivatives By using the new threshold function, a new adaptive shrinkage method is presented based on SURE rule. And a new denoising algorithm N-R with the combination of NeighShrinkage and relativity algorithm is presented based on the inter -intra mixed scale model. This two improved algorithm have the obvious superiority compared with the traditional methods. Moreover a simple non-gaussian noise model is setup and a non-gaussian denoising method is presented by using high order statistic, which gets a better result than the gaussian denoising method. At last we realize the hardware wavelet transform based on NiosII and FPGA and make it possible to simulate these developed algorithms on this hardware platform.