-In this paper we developed and manufactured a low-cost
pc-based hand geometry verification system that can be used in
various access control applications. The s/w and h/w were
optimized for the code, speed, performance, and for cost. A new
algorithm for the extraction of the hand geometry characteristic
features based on anatomical landmarks (fingertips and bottom of
valleys between fingers) were used in this system. These features
are extracted from the plan of the 2D_hand image captured by a
digital CCD camera and Infra Red Camera. These features were
used for the verification process. LCD, Keypad, and control
circuitry for Digital or IR camera were designed and integrated
within a PIII 550 system. Image processing operations such as
optimal thresholding, boundary extraction, thinning, and tracing
were used to get the anatomical landmarks (which are 45 points)
from the acquired grayscale images. Twenty-five distance and
area measures are extracted automatically based on the 45 points.
The features (Distances and areas) are the finger lengths, finger
areas, finger widths at different heights of the different fingers of
the right hand as well as the radius of the hand center. The
algorithm has been tested on 500 images successfully. Sensitivity,
specificity, and efficiency were calculated for the overall system
based on a digital or IR CCD cameras. The results were found to
be more than 99.7 % for the 500 images captured |
