Matching profiles of masked perpetrators: a pilot study
Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › fagfællebedømt
Standard
Matching profiles of masked perpetrators: a pilot study. / Lynnerup, Niels; Bojesen, Sophie; Kuhlman, Michael Bilde.
I: Medicine, Science and the Law, Bind 50, Nr. 4, 01.10.2010, s. 200-4.Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › fagfællebedømt
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - JOUR
T1 - Matching profiles of masked perpetrators: a pilot study
AU - Lynnerup, Niels
AU - Bojesen, Sophie
AU - Kuhlman, Michael Bilde
PY - 2010/10/1
Y1 - 2010/10/1
N2 - Seventeen male volunteers were photographed laterally with and without a tight-fitting balaclava. We then matched these photographs in blind trials. The matches were performed in two separate trials: as side-by-side comparisons, and by using superimposition. In both trials we graded the matches as good, possible and no fit. We found a very high true identification rate of 94% using superimposition, i.e. whether a match judged as good was in fact the correct match. The correct exclusion rate was also high, 94%, when using superimposition, i.e. when no match was found this was in fact correct. However, even though we were often correct in judging the best fit, we often had several possible matches for each case, which means that comparing profiles is not very selective. As such, using superimposition to compare the profile of a masked perpetrator with lateral photos of one or more suspects may indicate the possible matches, and perhaps even the best match, which may be helpful in police investigations, but it would not carry enough weight to be used as evidence per se. This study only focused on the profile. Future studies will use surface laser scans to analyse congruence between masked and unmasked subjects using the whole head.
AB - Seventeen male volunteers were photographed laterally with and without a tight-fitting balaclava. We then matched these photographs in blind trials. The matches were performed in two separate trials: as side-by-side comparisons, and by using superimposition. In both trials we graded the matches as good, possible and no fit. We found a very high true identification rate of 94% using superimposition, i.e. whether a match judged as good was in fact the correct match. The correct exclusion rate was also high, 94%, when using superimposition, i.e. when no match was found this was in fact correct. However, even though we were often correct in judging the best fit, we often had several possible matches for each case, which means that comparing profiles is not very selective. As such, using superimposition to compare the profile of a masked perpetrator with lateral photos of one or more suspects may indicate the possible matches, and perhaps even the best match, which may be helpful in police investigations, but it would not carry enough weight to be used as evidence per se. This study only focused on the profile. Future studies will use surface laser scans to analyse congruence between masked and unmasked subjects using the whole head.
KW - Adult
KW - Clothing
KW - Face
KW - Forensic Medicine
KW - Humans
KW - Image Processing, Computer-Assisted
KW - Male
KW - Photography
KW - Pilot Projects
KW - Software
M3 - Journal article
C2 - 21539287
VL - 50
SP - 200
EP - 204
JO - Medicine, Science and the Law
JF - Medicine, Science and the Law
SN - 0025-8024
IS - 4
ER -
ID: 33883644