A novel approach based on watershed algorithm to near infrared palm-dorsal vein image recognition is presented to avoid errors introduced in traditional vein image skeletonization stage.Skeletonization is performed directly on gray-scale vein image using watershed algorithm in our method.We use modified Hausdorff distance to evaluate the matching degree of vein patterns.Experiments were conducted on our own palm-dorsal vein image database and the results are satisfactory with 0.1% error equation rate (EER)

which indicates our method is a potentially usefull biometric.