According to the background suppression issue in infrared small target detection, a new method is put forward. The pure background of current viewing frame is estimated through the methods of distribution image and Bayes estimator based on the finite frame image. The images only containing targets and a few noise points are obtained through calculating each image and background images. The images are segmented and distinguished to suppress effectively the fluctuation background in images so that the error rate of post-processing is reduced. The effectiveness of the method is proved by experiments.