Dorsal hand vein pattern recognition system based on neural network

A new approach for authentication of individuals is presented in this paper. Use of the veins of the dorsal hand using infrared imaging and extraction of knuckle tip information is done. The knuckle tips are used as control points for obtaining the region which we are interested in that is the region of interest. The vein pattern is extracted to locate the branching points from which the intersection points are obtained. The distances and angles between each point and the other points is computed and a database of these features is created. Score generation and matching is done using the distance and angle combination between the intersection points. The system is then trained using neural network for better accuracy. The proposed system is contactless and low cost. This system produced satisfactory results and provides an user friendly means of verifying individuals.