A stage reduction speckle noise method of SAR images based on wavelet analysis

The interferential characteristic of Synthetic Aperture Radar (SAR) makes the SAR images produce a great deal of speckle noise, which strongly obstructs the interpretations and applications of SAR images. The traditional filter algorithms can restrain speckle noise efficiently, but edge information becomes blurry. Wavelet analysis decomposes a SAR image into a series of son images that represent different scale information. The low frequency images restrain speckle noise and the high frequency images which are processed properly with some thresholds can preserve good edge information, then they reconstruct the image. The experiments show that the stage reduction speckle noise method based on wavelet analysis can obtain good filter effect in SAR images.