Information Technology Reference
from the multiple samples and extracts feature from each template during Enrollment.
At the authentication stage, the inquiry fingerprint is matched with these templates
and then score level fusion or decision level fusion is used to give the final decision.
The second method [6,8] is to combine all the samples as a super template and extract
features from it. The third method [2,5,7] is to extract features from each sample first
and then combine the features together as one feature template. With the development
of computer hardware and software especially for storage and parallel processing
technique, storing multiple templates is feasible. Techniques on combining image
files or features mostly used in fingerprint recognition may not suitable for other bio-
metric recognition systems. Also, such techniques need deep comprehension of fea-
ture extraction and match details. In addition, it is difficult to combine too many raw
biometric samples or feature templates together. So it is helpful to select template
based on match scores after enrollment.
Template selection will improve the overall system's accuracy efficiently if enough
samples are captured in enrollment. However, it is not user-friendly to capture a num-
ber of fingerprints of the same finger at long intervals in the registration phase. Dur-
ing the authentication stage of a fingerprint recognition system, input fingerprints are
successively received and compared with the templates. If an input fingerprint is suc-
cessfully matched with a template, these two fingerprints are verified to originate
from the same finger. Therefore, input fingerprints can be used to update the matched
template. And score based template update is very important to improve the match
performance of biometric systems [3,12,13].
The rest of the paper is organized as follows. In Section 2 the model of templates
selection has been studied and the relation between sample number and template
number has been analyzed especially for the number of samples equal to 8. Based on
match scores, two algorithms have been described for template selection in section 3.
In section 4, two strategies have been proposed for template update. To study the
effectiveness of the proposed technique, section 5 gives the experimental results. The
last section summarizes the results of this work and provides future directions for
2 Multiple Templates Fusion
The match score is a measure of similarity between the template and the input biomet-
ric feature vectors. Combining different match scores with fusion strategy to give the
final decision is called score level fusion.
In daily life, people recognize each other usually by face and voice. In different
environments and conditions, we may often find that the similarity between two per-
sons is different. Although the probability that a large number of minutiae from im-
pressions of two different fingers will match is extremely small, fingerprint matchers
aim to find the “best”alignment . A fingerprint system compares the template and
input biometric traits in different angles and gives a match score for each angle. Fi-
nally, the greatest mark is chosen as the result which represents the greatest similarity
of the two fingerprints. We call this principle the greatest similarity principle.