Shearlet-based deconvolution under the framework of Bayesian

Shearlet transform which is based on a multiresolution analysis and built in the discrete framework provides efficient multiscale directional representations and yields approximately optimal representation properties. We employ the Laplacian model to describe the coefficients in shearlet domain, then in the MAP theory, a shearlet-based deblurring problem is equivalent to an optimization problem. Experimental results show the effectiveness of the algorithm corresponding to the optimization model and also show that the proposed deblurring method outperforms significantly than the existing prototype methods in Fourier domain and Wavelet domain in the perspective of both subjective vision and objective criteria.