A review of high-level multisensor fusion: approaches and applications

The potential advantages in multisensor fusion can be obtained more accurately, concerning feature that are impossible to perceive with individual sensors, as well as in less time, and at a lower cost. The characterization most commonly encountered in the rapidly growing multisensor fusion literature based on levels of detail in the information is that of the now well known triple low level (data level), medium level (feature level) and high level (decision level). The development of high-level multisensor fusion representations is very important, in the construction of advanced intelligent systems. The paper begins with a review on the fundamental principles about high-level multisensor fusion, together with some of the applications. Finally, we compare the decision algorithms in the high-level multisensor fusion.