Image Processing Reference
In-Depth Information
Sensor data—The revolution in miniaturization for computer systems has allowed
us to produce a myriad of sensors. The sensors can collect data about their en-
vironment (location, proximity, temperature, light, radiation, etc.), can analyze
this data, and can communicate between themselves. Collections of sensors can
produce very large streaming sets of data.
Video data—Video analytics are being used more and more to enhance the effec-
tiveness of the security in high-risk security operations. Content analysis, com-
bined with massive recording capabilities, is also being used as a powerful tool
for improving business processes and customer service. New techniques must be
developed to integrate this streaming data paradigm into the analyst's toolbox.
Whereas each of these categories can produce massive data streams containing
information that is applicable to a given problem domain, the grand challenge prob-
lem in the area of scalability is to use analytics to distill the relevant pieces of
information from these widely disparate information streams, and create an infor-
mation space containing relevant information that can be examined by analytical or
visual means to influence the exploration, hypothesis testing, discovery, and decision
making of the user. These systems need to provide mechanisms that can visualize the
connections between the relevant information in the information streams, and allow
the user to relate concepts, theories, and hypotheses to the data.
Several research directions present themselves as candidates to address these
scalability problems, classified as visual scalability, information scalability, software
scalability and information fusion.
28.2.1 Visual Scalability
Visual scalability [
1
] is the capability of visualization tools to effectively display
massive data sets, in terms of either the number or the dimension of individual data
elements. Factors affecting visual scalability include the quality of visual displays,
the visual metaphors used in the display of information, the techniques used to
interact with the visual representations, and the perception capabilities of the human
cognitive system. A critical area of research in visual scalability is in methods that
allow the user to change the visual representation of data.
28.2.2 Information Scalability
Information scalability implies the capability to extract relevant information from
massive data streams. Methods of data scalability includemethods to filter and reduce
the amount of data, techniques to represent the data in a multiresolution manner,
methods to abstract the data sets.
Search WWH ::
Custom Search