uwcse.location.core
Interface LocationSensor

All Known Implementing Classes:

AlienReaderFixed, AlienReaderMobile, AlienTagFixed, FixedSensorObject, MobileSensorObject

public interface LocationSensor

The interface LocationSensor should be implemented by objects which are sensors, that is, they generate Measurements.

Version:

$Revision: 1.2 $

Author:

Jeffrey Hightower

Method Summary
abstract  double	likelihood(Measurement meas, LocationReport sensorLocation, State objectState)           Determine the probabilistic likelihood [0...1] that this sensor generated this measurement about an object, given the State of this sensor and the State that object.

 

Method Detail

likelihood

public abstract double likelihood(Measurement meas,
                                  LocationReport sensorLocation,
                                  State objectState)

Determine the probabilistic likelihood [0...1] that this sensor generated this measurement about an object, given the State of this sensor and the State that object. More formally, concrete classes extending LocationSensor should compute the probability below. This quantity is referred to as the perceptual model in Bayesian filters.

P(Measurement|this sensor's State, measured object's State)

Note that providing an efficient implementation of this method is important because when using a particle filter, the likelihood is computed for all samples in the particle filter with each new measurement.

See Also:

Tracker.update(Measurement,LocationSensor,LocationReport)