Image Denoising Based On Sparse Representation And Wavelet Transform

Effective representation of image refers to using less data to express the interested target, means the ability of sparse representation. The research of sparse representation has became the popular topic in the field of image processing in recent years, Morphological component analysis is a decomposition method which is put forward based on a sparse representation of signals, There is the wide application prospect applied to express image information combine with other image processing method. The paper researches the following aspects based on the MCA sparse representation method:1, A systematic study has been made on the research of image denoising which based on wavelet transform method, and the numerical implementation of wavelet threshold denoising, modulus maximum denoising and related coefficient denoising, Contrasted and analysed the results of various methods of experiments.2, Combined with morphological component analysis method and the wavelet threshold denoising method to denoise the image. Specific include the Determine of MCA redundant dictionary, image decomposition, use wavelet transform to processing the sparse of the cartoon part. The experimental results showed that after MCA sparse representation of image denoising can also get a good denoising effect.3, Compared the results of different thresholds and threshold functions to denoise image which combined with MCA, and compared the results of without sparse representation of the denoising with sparse representation of the denoising, the experimental results showed that the sparse representation of the image was more conducive to image denoising.In conclusion, this paper specially researches the image denoising based on sparse representation, the experimental results showed that the sparse representation of images have gotten the better denoising effect.