DA3D: Fast and Data Adaptive Dual Domain Denoising

Nicola Pierazzo (1) , Martin Rais (2) , Jean-Michel Morel (1) , Gabriele Facciolo (3, 4, 5)

1 CMLA - Centre de Mathématiques et de Leurs Applications
2 Departament de Ciències Matemàtiques et Informàtica
3 LIGM - Laboratoire d'Informatique Gaspard-Monge
4 imagine [Marne-la-Vallée]
5 ENPC - École des Ponts ParisTech

This paper presents DA3D (Data Adaptive Dual Domain Denoising), a “last step denoising” method that takes as input a noisy image and as a guide the result of any state-of-the-art denoising algorithm. The method performs frequency domain shrinkage on shape and data-adaptive patches. Unlike other dual denoising methods, DA3D doesn’t process all the image samples, which allows it to use large patches (64 × 64 pixels). The shape and data-adaptive patches are dynamically selected, effectively concentrating the computations on areas with more details, thus accelerating the process considerably. DA3D also reduces the staircasing artifacts sometimes present in smooth parts of the guide images. The effectiveness of DA3D is confirmed by extensive experimentation. DA3D improves the result of almost all state-of-the-art methods, and this improvement requires little additional computation time.

Keywords: Image denoising; Patch-Based methods; Fourier shrinkage; Dual Denoising; Data Adaptive

Domains: Image Processing [eess.IV]; Signal and Image Processing; Modeling and Simulation; Computer Vision and Pattern Recognition [cs.CV]

Nicola Pierazzo, Martin Rais, Jean-Michel Morel, Gabriele Facciolo. DA3D: Fast and Data Adaptive Dual Domain Denoising. ICIP, 2015, Québec, Canada. ⟨hal-01240841⟩