Y. Li and C. Fu (PRC)
Pattern recognition, biometric, eyebrow recognition, feature string
Eyebrow is presented as a new biometric for personal identification. Based on feature strings which are extracted from eyebrow images, an eyebrow recognition method is described in detail and analyzed in experiments on a small eyebrow database taken from 32 persons. The results have shown that the presented method can reach the highest accuracies of 93.75% and 90.63% as well as the lowest accuracies of 87.50% and 81.25%, respectively, in two small-scale experiments. Since the recognition accuracies may be higher by improving the current method or introducing other ones, eyebrow is likely to work as a good biometric in the future.
Important Links:
Go Back
