The HandPunch uses the size and the shape of the human hand to verify a person's identity. It does not read fingerprints or the palm. The reader takes over 90 measurements of the fingers and part of the hand. The hand reader uses invisible light like the kind used in a television remote control, special optics and a CCD (chip) camera to capture a three-dimensional image of the hand. Using advanced microprocessor technology, the terminal converts the image to a template. It stores the template in database with the employee¡¯s ID number. To punch in, the employee enters his or her ID number at the terminal. HandPunch 2000 prompts the employee to place his or her hand. The terminal compares the hand on the platen with employee's unique template. If the images match, the terminal stores the punch information and uploads it to time and attendance software for processing.

The Recognition Systems, Inc. HandKey verifies a persons identity by looking at their unique hand geometry. Hand Geometry is the measurement of the size and shape of a hand. It includes length, width, thickness, and surface area of the four fingers and the hand. The thumb is not used. The HandKey must first capture a 3 dimensional image of the hand. To do this, an infrared light is used in conjunction with a 32,000 pixel CCD image array. The CCD image array is similar to those used in a common video camera. The infrared light is the same as what is used in a television remote control. The light illuminates the hand allowing the image to be captured by the CCD image array. The image is then used to determine the length of each finger, its width at several locations, the surface area near the knuckles, and the thickness of the fingers. In total, over 90 measurements defining the shape of the hand are derived from the image.

The measurements are then further processed to determine what are the unique characteristics of the hand. These unique features are then used to create the 9 byte template that is used for subsequent comparisons. An example of a unique feature might be the following: typically the middle finger is the longest finger. If the image provided showed that the middle finger was shorter than another finger, the HandKey would consider this fact a very unique aspect of the hand. This specific feature would be factored heavily into the resulting template for the person since this feature is very uncommon.

When an individual is enrolled in the HandKey, a template is created and stored in memory along with their ID number. The template becomes the reference template for future verification attempts. When the person uses the HandKey, they enter their ID number. The template associated with that ID number is brought into the HandKey¡¯s comparison memory. The user is prompted to place their hand and a template is created for that hand. This template is compared against the reference template to determine a ¡°goodness of fit¡± of the two templates. The result of the comparison is called the score. The greater the number of differences between the two templates, the greater the score. The opposite is also true. If the number of differences is low, so is the score. If the resulting score is lower than the set reject threshold, the user is verified. If the score exceeds the reject threshold, the user is rejected.