3D Multi-Object Tracking for Autonomous Vehicles Using Sensor Fusion
One challenging task for self-driving is ensuring safe navigation in high-dynamic environments like Urban regions. Locating objects such as cars and pedestrians and predicting their future movement provide adequate information to the autonomous vehicle to adjust its trajectory and speed accordingly. This task is called object tracking. In this research, we develop a persistent 3D multi-object tracking supported by a stereo camera, LiDAR, and IMU sensors that tackles the object occlusion problem.
