An Artificial Perceptual Model Approach - Artificial Human Sensors: Science and Applications - page 110

Figure 5.4. The execution of a low, medium and high level information fusion approach in
a hierarchical structure.
There is typically a basic connection between the three levels. Depending on
the outcome of the fusion system, after all it is of importance to meet the predefined
goal, with all the means at disposal.
Outgoing from a statement, a modified definition of fusion at different acces-
sible levels is proposed as:
“Sensor fusion/data fusion/feature fusion/information fusion is the process of combining
sensors/data/features/information in such a way that the result provides more added valued
information, compared to handling each source separately”.
Based on the motivation, a modified hierarchical structure from Fig. 5.3 con-
taining four different levels are illustrated below, in Fig. 5.5. The fourth structure
is a redundant fusion level, where the same parameter is measured and fused with
different type of sensors, e.g., all sensors are measuring a temperature parameter.
By adding an active perception process into a model, the performance will
ensure the possibility to obtain an increased performance in certain applications.