Fig.1

Initialized shape and control points"

Fig.2

Measurements received by sensors while tracking multiple targets"

Fig.3

Trajectories of the three targets"

Table 1

Initialization parameters for the targets"

Fig.4

Shape rendering of MLS-PMBM when estimating the ellipsoidal target"

Fig.5

Local magnification of MLS-PMBM when tracking the cube target"

Fig.6

Local magnification of MLS-PMBM when tracking the cross target"

Fig.7

Shape rendering of GGIW-PMBM when estimating the ellipsoidal target"

Fig.8

Trajectory diagram of GGIW-PMBM when tracking the cube target"

Fig.9

Trajectory diagram of GGIW-PMBM when tracking the cross target"

Table 2

Comparison of average IoU estimated by MLS-PMBM and GGIW-PMBM algorithms"

Table 3

Comparison of the time taken by single step arithmetic s"

Fig.10

Target number estimation in Experiment 1"

Fig.11

OSPA comparison of the two algorithms"

Fig.12

Result of the original point cloud visualization"

Table 4

Motion state of each stage of the target"

Fig.13

Local magnification of MLS-PMBM when tracking the vehicle target"

Fig.14

Local magnification of the tracking result when tracking the targets crossed"

Fig.15

Target number estimation in Experiment 2"

Fig.16

OSPA of this algorithm when tracking the car model"