Who we are

Babcock Training Academy

We deliver world-class training in complex systems to AUS/NZ defence, government and high tech markets

Office Address

Babcock Training Academy
Level 10 70 Franklin Street Adelaide  SA  5000

Contact Details

Training Project Support
T:   +61 (0)8 8440 9507
F:   +61 (0)8 8312 3227
This email address is being protected from spambots. You need JavaScript enabled to view it.


Tracking & Data Fusion

Course Description

Multiple target tracking is an integral component of modern surveillance systems for both military and civilian applications. This two-day short course delivered by Dr Branko Ristic, Dr Samuel Davey and Dr Jason Williams provides an introduction to the topic, spanning both classical methods, and modern developments such as random finite sets and track before detect.

The course commences with estimation methods for dynamical systems, covering the Kalman filter and its extensions for nonlinear and adaptive estimation. Subsequently, we examine methods for data association, building up to a description of the industry standard multiple hypothesis tracking (MHT) algorithm.

Many modern sensors allow the designer to dynamically adjust sensor parameters such as mode, look direction and platform path. Recent work in sensor management has provided a set of tools that exploit sensor flexibility to achieve improved performance. These methods will be outlined with the help of illustrative examples.

The second day of the course will study emerging topics in tracking, including random finite sets, a coherent theoretical framework for tracking problems which deeply integrates factors such as track initiation and termination; and track before detect, which provides improved performance by performing tracking directly on raw sensor data rather than using thresholded point detections.

Target Audience

The course is aimed at engineers and scientists working in radar, sensor networks, computer vision and other tracking applications.

Course Outline

Static estimation: Bayes rule, least squares estimation
Stochastic estimation: linear Gaussian systems, Kalman filter, non-linear extensions including extended
and unscented Kalman filters, and particle filters. Adaptive estimation methods including MMAE, IMM and
Data association: nearest neighbour methods, joint probabilistic data association, MHT and
implementation using multiple dimensional assignment. Implementation issues such as gating, hypothesis
management and track maintenance (initiation, termination).
Sensor management: optimisation criteria, greedy planning, open loop planning, closed loop
planning/dynamic programming, examples.
Tracking performance: prediction of error performance using theoretical lower bounds; evaluation using
MoEs, and the optimal-subpattern (OSPA) error.
Multiple sensor tracking and fusion: fusion architectures, information graphs, distributed fusion, track
association and fusion, registration of multiple sensors.
Bayesian tracking using random sets: Optimal multi-object Bayes filter, Bernoulli filter, PHD filter,
particle filter implementations with demonstrations.
Track before detect: TkBD for unthresholded radar data and video imagery. Numerical approximations
to optimal Bayes filtering, Histogram PMHT, ML-PDA.

Course Details

Course start
Course end
Individual price
Extra Information Contact us today about upcoming dates for this course or tailored courses to suit you and your team!
Course location
Please register your interest in this course by contacting Babcock on +61 (0)8 8440 1498

Register Here!

Registering for one of our courses is easy! Simply fill out one of the below Course Registration Forms and return via email to training@babcock.com.au

Course Registration Form - Individual Registration Form

Course Registration Form - Group Registration Form

Babcock Terms and Conditions


Want to attend a course?

If you would like to register your interest in a course that is not currently being run, please contact us.

GST Terms

All prices shown are Exclusive of GST. GST of 10% will be added to all invoices