Information Technology Reference
In-Depth Information
Chapter 6
E-UBIAS - Cognitive Systems for Image
and Biometric Data Analysis
E-UBIAS - Cognit ive Systems for Image and Biometric Data Analysis
E-UBIAS - Cognit ive Systems for Image and Biometric Data Analysis
UBIAS systems which conduct semantic analyses are also capable of biometric
identification using the following characteristics:
•
individual;
•
physical;
•
behavioural.
The main individual characteristics are determined for every image analysed.
However, if they concern a specific case, then they have to be identified in the
base in which they are defined as identifiers of that case. For instance, if what is
analysed is an image of the hand of someone with one finger much shorter in the
right hand, then the definition of the pattern for that person should contain a cor-
rectly defined pattern of that person's hand covering finger lengths, including the
unusual length of one of them.
However, if the item defined is the shape of the foot of a given individual for
whom the length of the first toe of the right foot is unusual, then this aspect has to
be included when defining the foot pattern. Otherwise it will turn out that the tra-
ditional patterns of the hand or the foot defined in the cognitive data analysis sys-
tem do not facilitate the correct semantic analysis due to the disparities between
the pattern and the specific individual.
A biometric analysis can also be carried out using the physical features of a
given individual. Basic physical traits that are used for biometric analyses include:
•
fingerprints;
•
iris;
•
retina;
•
face shape;
•
mouth shape;
•
ear shape;
•
hand shape;
•
palm lines;
•
geometry of the vascular system of a given body organ;
•
body temperature;
•
DNA;
•
smell;
•
timbre.
Biometric analyses are also carried out using behavioural traits, which now
mainly include:
•
voice analysis of the person (analysing their speech, singing or shouting);
•
handwriting analysis;
