Speckle noise reduction using multi-pass technique
Darwin Tjendera · The University of Queensland · 1999
Synthetic Aperture Radar processing is widely used in microwave remote sensing as a method of obtaining high resolution radar images. Although Synthetic Aperture Radar is a high resolution active microwave sensor, its processed data have nuisance noise called speckle noise which degrades image quality. The speckle noise, which has been referred to as multiplicative noise, arises from random variations in earth-surface roughness. This noise is conspicuous when the data are processed in small number of looks. Many speckle noise reduction techniques have been suggested in the past, with the Intensity Summation (IS) being the most commonly used. In this thesis, three methods of multiple-pass single- look processing for speckle reduction will be observed. These three methods are Intensity Summation (IS), Whitening Filter (WF) and Modified Whitening Filter (MWF). These methods combine multiple SAR images obtained by a satellite during multiple passes of the same region. Output is a single image that has the speckle noise reduced. Both simulated signals/images and satellite collected SAR images are considered. Two conclusions are reached. 1) The WF method is able to perform best in improving the signal to noise ratio for the satellite images. 2) The speckle noise in SAR is of a multiplicative nature.