ISAR motion compensation using entropy metrics

Radar imaging using an ISAR system requires the estimation of the targets kinematic parameters in order to be used for motion compensation. A focused and legible image requires this step. Some of the popular methods that perform this compensation are based on an entropy metric. This metric is relatively simple to use but requires a lot of computations (i.e. it can be quite slow) and it can also yield a false estimation due to local minima. In this work the authors introduced the use of a conditional entropy metric that helps with the problem of local minima and also propose the use of working with low resolution images to increase the speed of the estimation. Results are shown using real data.