It is available from various books by Yakoov Bar Shalom. I taught in many classes in connection with tracking, classfiication, and identification. can send slides
Don't just dive straight into JPDA. Start with the nearest-neighbor Kalman filter, then look at Probabilistic Data association (PDA). Make sure you understand how all the events are enumerated and their weights evaluated in PDA (for the single-target case) before you move on to JPDA (multi-target case) - otherwise you will just get confused. In PDA it helps if you appreciate that a single Gaussian is being fitted to a linear combination of Gaussians (a 'mixture') at the end of each update. The fit is done so that the 1st and 2nd moments of the single Gaussian and the many-component mixture are the same.
You can also find some applications and details in the book of Blackman :
Multiple-target Tracking with Radar Applications
- Badges
- Science topic
- Similar topics
- Psychology
- Cognitive Science
- Mind