Multiwavelet Image Denoising Method Research

With the widespread popularization of the image information, image has played an important role in many fields. However, in the course of processing image data, the quality of the image may be damaged because of many reasons. Take noise for example, as to the further management, it's necessary to denoise the image. In recent years, the multiwavelet which is from the wavelet gains great development, has been penetrating into each field and achieved some success in image denoising. Although there are many image denoising methods, using multiwavelet to denoise image still is a remarkable problem, and it's very significant to research in theory and application.In this thesis, I have done comprehensive research for the image denoising based on the multiwavelet transformation. The research contents are as follows : firstly, the multiwavelet theory is introduced in detail; secondly, traditional multiwavelet denoising method is researched and its develop potential is proved according to the experiment result; thirdly, on the basis of research in the relationship between multiwavelet coefficients, the image denoising method based on the intra-scale dependency of multiwavelet coefficients is proposed; fourthly, one new multilevel threshold is proposed based on classical shrinkage threshold proposed by Donoho, and we research the situation that image noise deviation is known and unknown; fifthly, the multiwavelet adaptive threshold denoising based on generation algorithm is proposed via combining generation algorithm and characteristic of image.As you know, multiwavelet theory has got rapid development. So, this thesis has done main introduction in multiwavelet transformation theory and its key technique which illustrated its superiority in theory, and Sa4 multiwavelet system used in our experiment in detail, which make a basic for the following work.According to the research of traditional multiwavelet denoising method, the feasibility and superiority of multiwavelet theory in image denoising has been proved in practice.It is well known that there are great relationships between multiwavelet coefficients. So we proposed multiwavelet coefficients dependency-based image denoising method according to the feature of intra-scale dependency of coefficients. The key point of this method is: after signal makes multiwavelet transformation, its coefficients will have strong relationship between each scale, but noise does not produce the obvious relationship. The results show that this algorithm is more effective...