E-UBIAS – Cognitive Systems for Image and Biometric Data Analysis - Advances in Cognitive Information Systems - page 89

the centres of gravity of individual hand bones;
•

•

the beginning and the end of every hand bone.
In both cases, hand geometry is unanimously identified by the length of indi-
vidual hand bones. For this notation, to identify the specific person, the records
containing model descriptions can be complemented with individual biometric
characteristics and the physical features, and in the case of very complex semantic
processes, behavioural traits can also be added. Complementing UBIAS systems
with biometric analysis processes has given birth to a new class of semantic analy-
sis systems called E-UBIAS ( Extended Understanding Based Image Analysis Sys-
tems ). This class of systems is designed for executing identification analyses, se-
mantic/biometric analyses and recognition analyses.
E-UBIAS systems carry this semantic analysis out as described below. The
identification system records the hand image for which the characteristic traits of
the hand are saved in an individual record. A diagram of characteristic traits as-
signed to a hand is illustrated in Figure 6.5, showing the introduced definitions of
particular measures and values assigned to hand geometry.

Fig. 6.5. Characteristic features of hand geometry

For the correctly described geometry of the hand, a record is defined in which
data is recorded in the form of a set:
B dl = { g ij , d ij , d ij-ij , g śri , d śri , g n }

where:
g ij - denotes the thickness of the bones of the i th finger and the j th phalanx for i
= {I, II, III, IV, V}, j = {1, 2, 3}
d ij - denotes the length of the bones of the i th finger and the j th phalanx for i =
{I, II, III, IV, V}, j = {1, 2, 3}