Hand geometry recognition technology is in broad use for access control as well as time and attendance applications. An individual places their hand on a reader, and their identity is verified based upon the location of a number of key points on their hand.



Traditional hand geometry system always uses pegs to fix the placement of the hand. Two main weaknesses of using pegs are that pegs will definitely deform the shape of the hand silhouette and users might place their hands incorrectly.



We have developed a fast and optimized algorithm for hand geometry recognition based on Neural Networks. Proposed approach does not require any particular hardware since extracted features are computed without assuming any fixed hand positioning and also a low-cost webcam can be used for image acquisition.