Particle Filter Tracking of Multiple Rolling Leukocytes in Vivo

In this paper, a novel multi-target tracking method is proposed, which is able to simultaneously track all the detected in vivo rolling leukocytes in the field of view, to stop tracking leukocytes moving out of the image, to identify new leukocytes in the frame and to initialize the tracking process. The proposed tracking method is based on the particle filter framework. The data augmentation algorithm is applied to reduce a high dimensional tracking problem to low dimensional one. The occlusion coefficient is introduced to handle occlusion between leukocytes. The target-measurement management is developed to accommodate the varying number of targets.