Tree structured filter banks for speckle reduction of SAR images

Speckle reduction and enhancement of synthetic aperture radar (SAR) images with simple tree structured filter banks are proposed. The filter banks, which are based only on IIR allpass filters, are computationally efficient and can be easily adapted to different sampling rates of the SAR image data. The proposed structure is useful for speckle reduction through its subband-images and the speckle reduction is obtained by thresholding the subband-image coefficients of the digitized SAR images. The thresholding is a nonlinear function which can be adapted for each of selected subband-images. Three thresholding methods, soft thresholding, hard thresholding and thresholding based on sigmiod function, are studied. The proposed tree structured filter bank methods show great promise for speckle removal and hence provide good detection performance for SAR based recognition.