This paper offers novel approaches to computing hand geometry measurements from frontal views of freely posed hands. These approaches offer advantages in hygiene, comfort, and reliability.
The ability to quickly compute hand geometry measurements from a freely posed hand offers advantages to biometric identification systems. Although hand geometry systems are not new, typical measurements of lengths and widths of fingers and palms require rigid placement of the hand against pegs. Slight deviations in hand position, finger stretch, or pressure can yield different measurements. The authors’ algorithm segments the hand from a known background under spotlights and locates feature points along the fingers and wrists. Given a database of 54 hand images, with three different images of the same hand of each subject, the developed approach uniquely identified a previously unseen hand with an overall accuracy of 92%. (Publisher abstract provided)
Downloads
Similar Publications
- Disaggregating the heterogeneity of DSM-5 posttraumatic stress symptoms and disorder among justice-involved adolescents
- Perceptions of the Future and Pregnancy Avoidance in the U.S.
- Physical and mental health upon reentry: The importance of social contact in prison for wellbeing during reintegration