作者: Michela Goffredo , John N. Carter , Mark S. Nixon
DOI: 10.1109/BTAS.2008.4699356
关键词:
摘要: We present a new method for front-view gait biometrics which uses single non-calibrated camera and extracts unique signatures from descriptors of silhouette's deformation. The proposed approach is particularly suitable identification by in the real world, where advantages completely unobtrusiveness, remoteness covertness biometric system preclude availability information CCTV images usually subjects an upper front-view. Tests on three different databases with walking towards have been performed. obtained results, mean CCR 96.3%, show that recognition individuals observed front can be achieved without any knowledge parameters. Moreover, has applied to directions results compared algorithms found literature. performance encouraging its appliance surveillance scenarios.