A Multisensory Monitoring and Interpretation Framework Based on the Model–View–Controller Paradigm - Foundations on Natural and Artificial Computation - page 443

Fig. 3. Framework layers
3.1 Layers of the Framework
The current section describes the framework levels (see Fig. 3). In first place,
the common model stands out. As aforesaid, this block is accessible from every
level. The levels are composed by MVC extended modules (see Fig. 1). Next, the
functionality of each level is described, although, in order to keep the framework
as generic as possible, no algorithm is associated. Of course, the proposed levels
are just a guideline to create the framework, but it is possible to include new
levels according to the application requirements.
Acquisition: This level directly interacts with the digital analog devices, mea-
suring from the physical world and adapting these measures to be usable by the
system. The measures are data from the sensors as well as data from other infor-
mation sources (disk, database, and so on). The acquisition level also performs
information preprocessing.
Sensor Fusion: This level is in charge of merging the sensor data to im-
prove the information quality (more complete and accurate). Fusion algorithms
may also operate with different spectrum images and are capable of introducing
knowledge on the domain.
Localization and Filtering: The third framework level is dedicated to isolate
the objects of interest contained in the input images. This level may hold a wide
range of methods, from the simplest one (a binarization applied to infrared (IR)
images [2]) to other more complex approaches yielding better results [8]. On
the other hand, this level also filters the spots corresponding to the objects of
interest with the aim of eliminating possible noise.