Comments on "Bayesian Approach to Extended Object and Cluster Tracking using Random Matrices".

This correspondence shows that the expression for the marginal density of the kinematic state in [1], which is valid only if the kinematic state is fully observable by the sensor, can be generalized. Our second comment concerns the direct integration of a scaling factor for the extent estimate. Thirdly, we show an equivalent update step based on sequential processing of single measurements (instead of the indirect parameters "mean measurement" and "measurement spread" in [1]). [ABSTRACT FROM PUBLISHER]