Hand geometry recognition systems are commonly available in two forms: full hand geometry systems that measure the entire hand, and finger geometry systems that measure only the index and middle fingers. Hand and finger geometry biometrics are automated measurements of hand and finger dimensions taken from a three dimensional image. A reader or camera captures up to 96 features of the hand such as the shape, width, length of fingers and knuckles, distance between joints, and the shape and thickness of the palm. This is similar to some types of facial recognition systems in that it examines the spatial geometry of the hand and fingers. Surface details such as fingerprints, lines, scars, dirt, and fingernails are ignored.
Now a mature technology, hand geometry was one of the first methods of biometric access control in the modern marketplace. Unlike other biometric access control systems, which have taken advantage of such technological breakthroughs as microchips or increased camera quality, the technology supporting hand geometry access control systems has not progressed significantly for many years. Hand geometry access control systems have been in use for almost 30 years at facilities that include nuclear power plants, welfare centers, immigration facilities, and day care centers.
Operation
Capturing a hand or finger geometry sample is straightforward . For hand geometry, the enrollee places their hand, palm down, in a detector containing a flat plate with five pins that guide the placement of the fingers. The detector registers the dimensions of the hand and fingers using an LED, a camera, and mirrors. Typically, three placements are required to enroll, and the enrollment template is a representation of the average of the measurements from the three placements.
Applications
Hand geometry is a relatively accurate technology, but it draws upon less data points than fingerprint, iris, or facial recognition. Hand geometry is quite useful in verification, such as matching a known user to that user’s enrollment template. The factor limiting hand and finger geometry to verification applications is that hands and fingers do not vary enough in dimensions to generate unique templates in large populations of users. Within the acceptance tolerances of the detector, each time a user seeks access, the data for more than one enrollee could generate a match. Therefore, an identification process, such as using a PIN, fingerprint, smart card reader, or some other device, should accompany hand geometry systems.
Hand geometry access control systems are especially useful in outdoor environments because the detectors are resistant to many environmental factors, such as ambient lighting or temperature changes. Hand geometry templates are small, making them portable and enabling the entire control system to be contained in a single detector unit. Since dirty or soiled hands do not affect hand geometry access control devices, these systems can be used at construction sites and foundries.
Performance Metrics
System operation is comparatively fast for access control. Access authorization typically takes place in two to three seconds. Throughput can be up to 10 users per minute.
Hand geometry FTE rates are low compared to some other biometric systems. FTE rates measure the likelihood that a user is incapable of enrolling in the system. Fingerprint recognition, for example, is prone to elevated FTE rates because of poor quality fingerprints, and facial recognition requires consistent lighting to enroll a user. Hand geometry does not suffer from such operational or environmental constraints. Hand geometry can have higher FMR rates making it more suitable for verification than identification. Since most users can use hand geometry for access control, few employees and visitors need to be processed outside the system.
Vulnerabilities
The primary vulnerability of hand/finger geometry systems is the probability that more than one person’s hand/finger geometry sample meets the threshold criteria for a match with a template is significantly greater than most other biometric systems. This also makes hand geometry more vulnerable to spoofing attacks.
Some situations involving a user can result in false rejections. Swelling or injury to the hand, the presence of bandages or jewelry, and changes in weight that affect the dimensions of the hand can result in false readings. Incorrect placement of the hand can also result in inaccurate measurement.
References :
- Access Control Technologies Handbook
see also :