Low-Cost Hand Vein Pattern Recognition
Marion Distler, Sebastian Hoppe Nesgaard Jensen, Niels G. Myrtue, Claire Petitimbert, Kamal Nasrollahi, Thomas Baltzer Moeslund · VBN Forskningsportal (Aalborg Universitet) · 2011
Identification based on hand vein pattern is an interesting branch of biometric recognition that is enjoying increasingly attention among the researchers in the recent decade. However, despite the fact the most of the developed commercial systems for this purpose are expensive, their false acceptance ratios are not zero. Therefore, using them for identification purposes can be crucial if they mistakenly accept an imposter (even with a very low probability). The current work proposes a very low-cost hand vein pattern recognition system using a simple modified webcam. The system introduces a blob removal algorithm for enhancing the results of the segmentation and uses a modified version of Hausdorff distance for feature matching for the recognition purposes. Experimental results show that the system can achieve a zero false acceptance ratio while keeping the true acceptance rate in an acceptable level. KeywordsPattern Recognition; Hand Vein; Biometrics; Infrared Imaging.