This one-day course provides hands-on experience with developing and testing localization and tracking algorithms. Examples and exercises demonstrate the use of appropriate MATLAB® and Sensor Fusion and Tracking Toolbox™ functionality.
- Localization for orientation and position
- Scene generation and sensor detection import
- Filters and motion models
- Data association
- Multi-object trackers
Day 1 of 1
Localization for Orientation and Position
Objective: Fuse IMU and GPS sensor data to estimate position and orientation.
- Model measurements from accelerometers, gyroscopes, magnetometers, and GPS
- Fuse sensor data to estimate the pose in terms of position, velocity, and orientation
- Visualize the pose estimation and plot platforms and trajectories
Scene Generation and Detection Import
Objective: Import and process detections or generate scenarios used in multi-object trackers.
- Preprocess and package collected sensor detections
- Create a tracking scenario with multiple sensors and platforms
- Define waypoint or kinematic trajectories
- Customize sensor parameters
- Generate detections used in sensor fusion algorithms
Filters and Motion Models
Objective: Select and tune filters and motion models based on tracking requirements.
- Evaluate filters against scenario requirements
- Compare and contrast different motion models
- Configure an Interacting Multiple Model (IMM) filter to track different maneuvers
Objective: Determine the appropriate data association method for different tracking situations.
- Select from among Global Nearest Neighbor (GNN), Joint Probabilistic Data Association (JPDA), Track-Oriented Multiple Hypothesis (TOMHT), and other data association methods
- Determine how multiple detections are assigned to different tracks
Objective: Create multi-object trackers to fuse information from multiple sensors such as vision, radar, and lidar.
- Configure trackers and parameters
- Perform track association and management
- Visualize the tracked objects
Appendix A: Trackers for Passive Sensors
Objective: Create multi-object trackers and fusion systems that receive angle-only or range-only measurements from passive sensor systems.
- Triangulate multiple line-of-sight detections
- Perform static fusion of passive synchronous sensor detections
- Track with range-only measurements
- Track with angle-only measurements