Poisson Noise Removal with Total Variation Regularization and Local Fidelity

In this paper, we propose two methods to choose the fidelity parameter in total variation based denoising model for Poisson noise. Firstly, we derive a scheme to choose the scalar parameter automatically. Secondly, we propose to use a local fidelity term with spatial-varying parameters which automatically controls the extent of denoising according to image contents. Experiments with simulated data demonstrate that the proposed algorithms are effective.