New median filtering method based on statistic analysis

In order to improve the filtering effect of median filtering, a new method was put forward taking the local variation and local median value into account. The corrupted pixels were detected at first based on the discretion of the data and then filtered based on statistic characteristics. After computing the speckle indexes, peak signal-to-noise ratio and normalized mean square error of the processed images, their ability of suppressing speckle noise was compared. The results indicate that the new filtering method does have advantages.