Kullback leibler estimation of distribution parameters for SAR data

It is essential for SAR automatic target recognition (ATR) to estimate the parameters of statistical modes in SAR images. Most studies have relied on the conventional methods, such as moments technique and maximum likelihood estimation (MLE) method. In this paper, we apply a Kullback Leibler (KL) method to parameters estimation for SAR data. The ant colony algorithm (ACA) is extended to search for the optimal parameters. Experiment results show that the KL method performs better than MLE method in all cases tested.