A variational enhancement method for infrared images was proposed based on total variation(TV) and contrast.Considering that the infrared images generally are much noisy,the proposed method restricts the noise to compute the target gradient based on the existing variational enhancement scheme.In addition,the object function includes an extra term representing total variations,which gives a further suppression to the noise.The histogram-based algorithms are likely to merge the gray levels which are similar in intensity and have low distribution,thus some useful details will be decreased.The proposed variational method does not have such problems as histogram method,and is more effective to enhance infrared images.