DocumentCode
605562
Title
Exploring the uses of fingerprint patterns and ridge-counts in biographical associations
Author
Herdegen, D.W. ; Loew, M.H.
Author_Institution
MITRE Corp., McLean, VA, USA
fYear
2012
fDate
9-11 Oct. 2012
Firstpage
1
Lastpage
5
Abstract
Fingerprint characteristics describe people and groups. A person can be described approximately by fingerprint patterns and ridge-counts, and described uniquely by fingerprint minutiae. A group can be described by overall pattern frequencies and average ridge-counts (per-person or per-finger). When a group comprises related persons, the genetic basis of fingerprints produces common qualitative and quantitative fingerprint characteristics within the group. Those common characteristics allow for differentiation between endogamous (related by marriage) groups; forensic anthropology is replete with examples detailing the dermatoglyphic differences between endogamous groups based on race, religion, geography, or caste. This paper examines the degree of differentiation between discrete endogamous groups and explores the ability to associate an individual to a group by comparing individual-to-group fingerprint characteristics. Data from dermatoglyphic anthropologic studies cited herein are used to illustrate the ability to differentiate groups based on fingerprint pattern and/or ridge-counts. In some instances, the degree of differentiation between groups suggests that it is possible to identify associations of individuals to a group. A case study is presented that illustrates the association of a person to one of two endogamous groups by comparing individual per-finger patterns and ridge-counts to composited group pattern and ridge-count information. In this case, approximately 80 percent association accuracy was achieved using a decision tree classifier. The success achieved in this limited case suggests further study in associating persons to groups using fingerprint characteristics.
Keywords
biographies; cultural aspects; decision trees; edge detection; fingerprint identification; image classification; image forensics; average ridge-counts; biographical associations; decision tree classifier; dermatoglyphic anthropologic studies; dermatoglyphic differences; discrete endogamous groups; fingerprint characteristics; fingerprint minutiae; fingerprint patterns; forensic anthropology; overall pattern frequencies; qualitative fingerprint characteristics; quantitative fingerprint characteristics; Fingerprints; classification;
fLanguage
English
Publisher
ieee
Conference_Titel
Applied Imagery Pattern Recognition Workshop (AIPR), 2012 IEEE
Conference_Location
Washington, DC
ISSN
1550-5219
Print_ISBN
978-1-4673-4558-3
Type
conf
DOI
10.1109/AIPR.2012.6528220
Filename
6528220
Link To Document