Research On Wavelet Transform's Application To Image De-noising And Compression

Wavelet transform has been broadly used in digital signal processing applications as an important time-frequency transform. It has a perfect performance on both local time and frequency domain, so it is an effective analytical method. This dissertation put emphasis on wavelet transform's application to image de-noising and compression.First of all know the wavelet transformation, the forming process of the mathematical properties and in image processing area applications. For further research on wavelet transform and realizes image noise reduction and compression provides theoretical basis. Clear the image noise reduction and compression of background and practical meaning, and image noise reduction and compression method are reviewed. Around three classic wavelet de-noising methods, lists the wavelet threshold noise reduction method of the deficiencies, expounds the bayesian noise reduction method. Simulation results show that the bayesian noise reduction method improves the noise performance, including the noise image noise reduction effect improve. Deeply discuss the embedded zerotree wavelet coding algorithm, combined with soft threshold function and with the quantitative technology, both died in bayesian atrophy threshold and embedded zerotree wavelet coding algorithm is proposed on the basis of image with noise reduction and compression coding scheme, and of the image noise reduction and compression effectively combined with wavelet transform and realize image noise reduction and compression coding, and in the same compression ratio under the conditions of the EZW is direct compression method, JPEG compression method and the method of this paper, through the simulation experiment shows the method of this paper can get better reconstruction image quality.