Nose biometrics verification using linear object technique

摘要

Biometrics technology is evolving. Thus, researches have been done and implemented on many biometrics systems such as hand geometric, palm print, iris, voice gait. All these are sensitive to a number of factors such as background noise, high cost of deployment, inappropriate for large-scale identification, and finally gait-based system by the choice of footwear, nature of clothing, affliction of the legs, walking surface. However, the use of other characteristic feature such as the nose which is not bogged down with the aforementioned factors has been largely unexplored. In this work, geometry is employed for nose identification; the nose alae are measured. The curvature points of the nose alae are selected as the nasal feature using linear object class and triangulation technique. Our experiment was carried out with 100 dataset, the result showed 87% recognition; this shows that nose biometrics, if well explored could be a valuable biometrics traits.