Linear multitarget integrated track splitting for multiple detection target tracking

Abstract

This paper deals with multiple detection target tracking. We propose a novel algorithm to cover the many-to-one measurements-to-track association where the linear multitarget integrated track splitting algorithm is extended to accommodate for multiple detection target tracking using measurement partition method. After measurement-selection, measurement cells, where one measurement cell may contain one or many measurements treated as possible target detections, are generated by measurement partition method. Then the data associations and track components generation for each track are based on these measurement cells. The data association complexity of the proposed algorithm is linear in the number of tracks and measurement cells. Numerical examples are given to illustrate the benefit.