Adaptive average BM3D filter for reconstruction of images with combined noise

A method for reconstruction of images corrupted by mixed, salt-and-pepper and Gaussian, noise is proposed. The adaptive average BM3D filter is a combination of two algorithms for the reconstruction of images corrupted with mixed noise. The two-stage adaptive average filter showed good results in the denoising of images corrupted by a salt-and-pepper noise, whereas the BM3D algorithm showed good results in denoising images corrupted by a Gaussian noise. The algorithm was compared to a state-of-the-art algorithm.