USING HIDDEN MARKOV MODEL TO DETERMINE OBJECTIVE IN MULTI-TARGET TRACKING

113 views

Abstract

Multi-Target Tracking (MTT) has many practical applications, especially in national defense and security. The research results published so far mainly use Bayesian Sequential Estimation (BSE) to update the status and build algorithms to follow the orbits of targets. Those algorithms are all non-trivial algorithms because they are associated with very complex random models. The two most important issues for MTT are: determining the number of targets available at each time and determining their movement trajectories.

The published trajectory algorithms have difficulty identifying targets in case the new target appears at the time of the current observation. In this paper, we present a research result to solve the problem of determining the number of targets in MTT at any time, which can be overcome difficulties as mentioned earlier with techniques using hidden Markov Modeling Tool (Hidden Markov Model - HMM). The technique of using the HMMs tool in solving MTT problems, in the published results, no work has been mentioned.