Abstract: Object motion or camera shake during exposure often leads to image blur.This paper proposes an algorithm for removing motion blur in images.Based on the observation that clear images often have step edges,we develop a new type of edge region constraints by analyzing the relationship between the step-edge region and the motion kernel.With this type of constraints,the motion kernel is efficiently estimated by using the RANSAC algorithm.Extensive comparative experiments are conducted,and verify that our algorithm is simple,highly accurate,and is capable of restoring higher quality images than traditional methods.