Wavelet-based two-level methods for image restoration

Wavelet-based multilevel approach is a promising method for image restoration problems. In this paper, a method for image restoration based on combining the symmlets family with two-level technique is proposed. A key feature is that the symmlets have many flexible attributes that lead to better stability and less artifacts. Computational experiments show that symmlets-based two-level method yields smaller relative errors and more stable results.

► A method for image restoration based on combining the symmlets family with two-level technique is proposed.
► Regularization methods are exploited in the process of two-level method.
► The symmlets have many flexible attributes that lead to better stability and less artifacts.
► The proposed method based on combining symmlets performs better than the method based on Haar wavelet.
► Symmlets-based two-level method yields smaller relative errors and more stable results.