Background estimation is one of the most effective method in clutter suppression and image enhancement.It is widely used in infrared target detection.In this paper,we propose a new background estimation method with improved morphological filter.At first we compute the local image entropy and establish the relationship between the structural elements and local entropy.We choose the shape of structural elements based on target characteristics analysis.Then we do the background estimation by using morphological filter for clutter suppression and image enhancement.Simulation results show that the algorithm is effective.