Data restoration based on Gaussian noisy and motion-blurred snapshots in multimedia big data

In the era of big data, the complexity of data analysis and information extraction has grown dramatically. In this paper, we firstly analyze the disadvantages of TV deblurring model and TV-\( {\mathit{\mathsf{L}}}^{\mathsf{1}} \)denoising model. Afterwards, we propose an alternating iterative optimization (AIO) algorithm to data restoration based on two noisy and blurred degraded snapshots. Next, we prove theoretically the existence, uniqueness of the minimal solutions and convergence of the AIO algorithm, which are the most important part of the paper. In the end, we show our discrete numerical algorithm and some experimental results. Through these experiments, we can see that the AIO algorithm proposed by us works effectively very well.