An application of exponential smoothing technique in image restoration

In this study, a new smoothing technique based on the exponential function is proposed. A smoothing algorithm for minimizing the total variation model of energy is designed by using a new exponential smoothing function. A smoothing algorithm is combined with gradient descent algorithms with different line search techniques. The convergence of the algorithms have been theoretically proven. The algorithms are applied to some test problems and numerical results are presented. The obtained experimental results are compared with the results obtained by using hyperbolic smoothing functions and standard total variation (TV) approaches. The numerical results show that the proposed algorithms are promising and have remarkable advantages over competing algorithms.