We model the problem of video denoising as an optimization problem. Inter-frame information is exploited to provide lowrankcharacteristics of an input video stream. The noise-free video frames are extracted as the low rank components of a convexoptimization model. We use robust principle component analysis (RPCA) to extract away the noise from the input video streams.The fast version of RPCA, fast RPCA (FRPCA), is used for realtime denoising. A noisy video stream is considered as a 3D noisysignal which is a summation of a 3D noise-free signal and sparse additive noise. Exact augmented Lagrangian multipliers (EALM)method is used to solve the model for the low-rank terms, which represent the noise-free frames. The proposed approach has severaladvantages over existing video denoising approaches. It is model-independent, i.e. it does not require shape, appearance, or speedmodels. Also, it does not need prior information about the acquisition environment. Three different sets of data were used forevaluation: synthetic data for simulation experiments to provide quantitative results, real static videos, and real dynamic videos. Theevaluation results proved the effectiveness of the proposed approach when compared to existing approaches.
展开▼
