Artificial intelligence (2nd ed.): structures and strategies for complex problem-solving
Artificial intelligence (2nd ed.): structures and strategies for complex problem-solving
Learning in graphical models
Introduction to Bayesian Networks
Introduction to Bayesian Networks
| Hi-index | 0.00 |
In this paper we examine probabilistically the incorporation of contextual information into an automatic target recognition system. In particular, we attempt to recognise multiple military targets, given measurements on the targets, knowledge of the likely groups of targets and measurements on the terrain in which the targets lie. This allows us to take into account such factors as clustering of targets, preference to hiding next to cover at the extremities of fields and ability to traverse different types of terrain. Bayesian networks are used to formulate the uncertain causal relationships underlying such a scheme. Results for a simulated example, when compared to the use of independent Bayesian classifiers, show improved performance in recognising both groups of targets and individual targets.