Robust video denoising by low-rank decomposition and modeling noises with mixture of Gaussian

Abstract

This paper introduces a new approach for video denoising. Based on the idea of patch based low rank matrix completion, we improve the method by modeling noises with Mixture of Gaussians (MoG). By utilizing a series of different gaussian distributions to fit the representation of video noises without any assumptions on the statistical properties, the parameters of MoG are learned from video data automatically. It can deal with the fact that for most of the time, the real distribution of noises appeared in videos are unknown so that traditional methods do not work well without any priori knowledge. After the model and algorithm statements, we provide a group of experiments on real videos for comparisons with the state-of-art video denoising algorithm, which demonstrates the effectiveness and advantage of our approach.

Robust video denoising with sparse and dense noise modelings

“…It is an extension to the research presented in our previous conference paper [8]. The main differences are the inclusion of the additional l 1 norm, and a thorough experimental evaluation.…”

“…The number of Gaussian components (denoted by N ) needs to be adjusted according to the variance of the Gaussian distribution, which can refer to our previous paper [8]. The low rank matrix components U and V are formulated using…”

“…The competitive methods include cVBM3D [1], MoG [8], LRMC [3] and WNNM [9]. We synthesized mixed noises with two categories: continuous and sparse.…”