Department of Computer Science and Electrical Engineering

Weighted Sum Speckle Reduction of Multi-Pass Images

Peter Sohner

Abstract:
This thesis presents a weighted sum speckle reduction of multi-pass single-look synthetic aperture radar (SAR) images.  Three different filtering techniques will be applied to both simulated and SAR image data.  The methods used are Intensity Summation (IS), Weighted Intensity Summation (WIS), and Autoregressive (AR).

The mathematical matrix program called Matlab is used to apply these methods to the SAR images obtained by the ERS-1 satellite over Bonn in Germany.  It is also used to simulate images and determine the optimal method of speckle reduction by measuring the equivalent number of looks (ENL).

The results showed that the AR method was optimal for the simulated images that had additive noise and the WIS method was optimal for the SAR images, which contained multiplicative noise.

It can be concluded that a high reduction of speckle can be achieved using multi-pass single-look images and by applying the WIS method to them.  However, a combination of several techniques will produce the best speckle reduction.
 

Complete thesis:
thesis.pdf

Conference Paper:
 conference_paper.pdf

About the Author