Maintaining track identity for a fragmenting target
Anthony Gardner, Malcolm P. Rollason, Nick Everett, David J. Salmond · 2006
A multiple hypothesis track splitting filter has been developed for maintaining track on a target as it sheds extraneous objects. The filter includes full modelling of merged measurements from closely spaced objects and the transition to resolution of individual objects. This method is based on Kalman filtering of measurement streams (or tracks) defined by frame-to- frame associations which are assumed to be known.