Abstract
A video super-resolution (SR) reconstruction algorithm based on total variation (TV) regularization is proposed in this paper. With the analysis of image degradation, the image formation model is introduced. Total variation method is chosen to regularize the ill-posed problem and reconstruct the SR image. Instead of computing the complicated nonlinear partial differential equation (PDE) of the TV regularization in SR reconstruction, the quadratic upper bound function minimization method is used to efficiently solve the optimization. Results of the experiments show that the proposed algorithm has better subjective visual effect, and better peak signal to noise ratio, structural similarity and convergence, which confirms the effectiveness of the method.
Keywords : Super-resolution; Video reconstruction; Total variation regularization; Quadratic upper boundfun
