Restoring images corruptedby noise and blur is a burgeoning subject in image processing and, despite the large numberof proposed restoration algorithms, the effort to bring about some improvement is always ofgreat interest.The definition of fractional derivatives in recentyears has created a powerful tool for this purpose.In the present paper, using fractional-order totalvariation and framelet transform,the nonconvex model forimage restoration with impulse noise problem is improved.Then by alternating direction method of multipliers (ADMM) andprimal-dual problem, the proposed model is solved. The convergence of the proposedalgorithm is studiedand the proposed algorithm is evaluated using different types of tests.The output results show the effectiveness of the proposed method.