13-09-2013, 04:28 PM
Dental Biometrics: Human Identification Based On Dental Work Information
Dental Biometrics:.pdf (Size: 524.67 KB / Downloads: 49)
Abstract
Dental biometrics is used in forensic dentistry to
identify or verify persons based on their dental
radiographs. This paper presents a method for human
identification based on dental work information. The
proposed method works with three main processing
steps: segmentation (feature extraction), creation of a
dental code, and matching. In the segmentation step,
seed points of the dental works are detected by
thresholding. The final segmentation is obtained with a
snake (active contour) algorithm. The dental code is
defined from the position (upper or lower), the size of
the dental works, and distance between neighboring
dental works. The matching stage is performed with
the Edit distance (Levenshtein distance). The costs for
the insertion, deletion and substitution operations were
adapted to make the matching algorithm more
sensitive. The method was tested on a database
including 68 dental radiographs and the results are
encouraging.
Introduction
Biometrics is the science and technology of
identification, i.e. establishing the identity of an
individual, by measuring the subject’s physical or
behavioral traits. The term is derived from the Greek
words "bios" for life and "metron" for to measure [1].
The method of dental biometrics is used in forensic
medicine (forensic dentistry) to identify persons by
matching post-mortem radiographs (acquired after a
person is deceased), with ante-mortem radiographs
(acquired before a person is deceased) in a database, but
can also be used to match two ante-mortem or two post-
mortem radiographs (see Fig. 1). In some cases (plane
crashes, fire accidents, etc.) biometric features such as
faces or fingerprints are destroyed and it is not possible to
work with conventional identification methods like
fingerprint or face recognition.
Distance between two dental works
To make the matching algorithm more sensitive,
also the distance (amount of pixels) between two
neighboring DWs is included into the DC. The
distance is defined by the amount of pixels between
the center of mass points of the two DWs. The distance
of the leftmost DW (d1) is set to zero to make the
algorithm more stable against small deviations in the
manually registration of the DRs (see Fig. 10). The
value for the distance is given in percentage of the total
width of the DR, which is always 1000 pixels.
RESULT
A database including 68 DRs was used in the
experiments to evaluate the proposed dental biometric
method: a pair of DRs for 22 subjects (44 radiographs)
plus a single DR for other 24 subjects. For the 22
subjects with two radiographs, their oldest DRs were
considered as AM (ante-mortem) radiographs and their
newest DRs were considered as PM (post-mortem)
radiographs. For the 24 subjects with only one
radiograph, the DRs were considered as AM
radiographs. The images were manually registrated to
obtain comparable conditions. In cases of over- or
under segmentation of the DWs, the segmentation
result had to be corrected manually (see Fig. 12). Also,
if DWs are not detected by thresholding, a ROI has to
be selected manually in the DR to perform local
thresholding. Segmentation results of two distinct DRs
and their corresponding DCs are shown in Fig. 13.
Conclusion and Future Work
Dental biometrics is used to identify individuals in
the forensic medicine. This paper presents a dental
biometric identification method based only on dental
work information extracted out of panoramic DRs.
For the database used in the experiments, the results
obtained were good (11% of EER and 86% of
accuracy for top-1 retrieval), especially considering
that only dental work information is considered by the
method.
Because the amount of images in the database is
low, it is not possible to make a clear statement about
the performance and effectiveness of the proposed
method. In future work, the proposed method will be
assessed on a larger database. Also, in future work, the
proposed method will be fused to other methods based
on different tooth features.