Referring to the problem of slow convergence rate of current Total Variation (TV) image restoration algorithm, a class of fast TV image restoration algorithms was proposed by using multi-step iteration method. And, two kinds of multi-step TV restoration algorithms were proposed: multi-step iteration shrinkage/threshold algorithm and multi-step iteration re-weighted shrinkage algorithm. Furthermore, to overcome the shortcoming of weight parameters estimation at each iteration step, a kind of fixed weigh parameters multi-step TV restoration algorithm was given. Compared with the traditional single-step TV restoration algorithms, the computational complexity of the proposed algorithms was not added in each iteration step and the convergence could be achieved by less iteration numbers. Experimental results demonstrate that the convergence performance of our proposed algorithms is much higher than that of traditional single-step iteration restoration algorithms by recovering different type degraded images, so our algorithms are efficient.