Fusion of Non-Simultaneous Elements of Information: Temporal Fusion - Information Fusion in Signal and Image Processing - page 245

11.8. Conclusion
Whenever trying to track the evolution of variables in time, it is essential to study
the temporal constraints:
- the need for information set by the application (frequency of data availability);
- the temporal features of the sensors and of the associated data processes (fre-
quency, delay);
- the temporal characteristics related to the sensors.
After listing all of these, the performances of the sensors can turn out to be good
enough as to not require the use of mechanisms specific to temporal fusion (the acqui-
sition frequency is high enough, the delays are negligible). If this is not the case, a
system has to be implemented for assigning dates to the data and evolutionary models
are needed in order to achieve temporal registration.
There are few studies that have directly focused on temporal fusion, or more gen-
erally on time management in data fusion applications. This aspect appears mostly in
military applications for target tracking and in robotics. In most cases, time manage-
ment is limited to the use of the Kalman filter, often in its extended version (lineariza-
tion around an operating point). The quality of the fusion results could probably be
significantly improved if, in critical applications, the temporal aspect was not plainly
ignored.
11.9. Bibliography
[ABI 92] A BIDI M.A., G ONZALEZ R.C., Data Fusion in Robotics and Machine Intelligence ,
Academic Press, 1992.
[ALL 01] A LLOUCHE M.K., “Classification of Temporal Information in Situation Analysis”,
ISIF, Fusion 2001 , vol. 2, International Conference on Information Fusion, p. ThC3 11-18,
Montreal, Quebec, Canada, 2001.
[ANK 01] A NKEN C.S., B URNS C.L., “Theater Ballistic Missile (TBM) Reasoner: A Knowl-
edge Base Decision Aid for Time Critical Targeting Situation Assesment”, ISIF, Fusion
2001 , vol. 2, International Conference on Information Fusion, p. FrA2 13-18, Montreal,
Quebec, Canada, 2001.
[APP 98] A PPRIOU A., “ Uncertain Data Aggregation in Classification and Tracking Pro-
cesses ”, Aggregation and Fusion of Imperfect Information chapter, Physica-Verlag, 1998.
[BAR 88] B AR -S HALOM Y. , F ORTMANN T., Tracking and Data Association , vol. 179, Aca-
demic Press, 1988.
[ELE 96] E LETER B., R OMBAUT M., “Intelligent Mapping of the ProLab2 Vehicle Dynamic
Environment”, Mathematics and Computer in Simulation , vol. 41, no. 3-4, p. 329-336,
1996.