*@book{Stone99Bayesian,*

Year = {1999},

Title = {Bayesian Multiple Target Tracking },

Isbn = {1580530249 },

Booktitle = {Bayesian Multiple Target Tracking },

Publisher = {Artech House},

Author = {Lawrence D. Stone, and Carl A. Barlow, and Thomas L. Corwin}}

Year = {1999},

Title = {Bayesian Multiple Target Tracking },

Isbn = {1580530249 },

Booktitle = {Bayesian Multiple Target Tracking },

Publisher = {Artech House},

Author = {Lawrence D. Stone, and Carl A. Barlow, and Thomas L. Corwin}}

Likelihood functions replace and generalize the notion of contact used in linear Gaussian trakers. For this reason, likelihood functions play a key role in Bayesian tracking. The likelihood function L for the random variable X and observation Y=y is defined to be: $L(y|x) = Pr{Y=y | X=x} for x\in S$. Baye' theory: $p(x|y) = \frac{L(y|x)p(x)}{\int L(y|x)p(x)dx}$. The distribution given by p is called the prior distribution and the distribution given by $p(.|y)$ is called the posterior distribution on X given the observation Y=y. There are several typical likelihood functions such as: Gaussian contact model, Gaussian bearing-error model, Single-plus-noise model.

Other links from Mathworld

## No comments:

## Post a Comment