Maintaining track continuity in GMPHD filter

Abstract

The data association between objects and measurements is a challenging task in multiple-object tracking because of computationally expensive. This challenge can be overcame by the probability hypothesis density (PHD) filter. Recently, the Gaussian mixture probability hypothesis density (GMPHD) filter has been proposed as a closed-form of the PHD filter. However, the GMPHD filter does not include track continuity during the period of tracking. In this paper, we present a method for maintaining the continuity of state estimates of objects in the GMPHD filter. The set of labels from Gaussian components is used to create hypotheses for label association process and the Hungarian algorithm is applied to search for the best hypothesis association. The results show that the method is robust and efficient. © 2007 IEEE.