Nonlinear Edge-Preserving Smoothing of Synthetic Aperture Radar Images

Synthetic aperture radar (SAR) images are subject to prominent speckle noise, which is generally considered a purely multiplicative noise process. One interesting property of this multiplicative noise is that the ratio of the standard deviation to the signal value, the “coefficient of variation,” is theoretically constant at every point in a SAR image. We use this property in conjunction with a new nonlinear filter structure based on mathematical morphology, the value-and-criterion structure, to design a filter that removes speckle noise from SAR images without blurring edges. First, the sample coefficient of variation at each point in the image is computed. In areas where there are changes in the signal, the sample coefficient of variation will be greater than the expected theoretical value. By using the new filter structure, a low-pass filter to remove speckle noise can be directed to operate only over regions where the coefficient of variation is close to the expected value. These regions are less likely to contain significant features or edges which would be distorted by low-pass filtering. We demonstrate the effectiveness of this new filtering method by comparing it to established speckle noise removal techniques on both phantom images with simulated speckle noise and real SAR images.