摘要
针对基于人体特征展开的身份认证还停留在静止认证的阶段,采用人体分割和状态预测的方法展开基于运动状态人体的身份认证.首先从静止状态下的人体分割方法进行研究,确定了研究的基础骨架;其次,对经典的HMM理论进行适当拓展使其适用于运动状态下的人体状态预测研究;最后,结合前述研究结果,完成了运动状态下的身份认证研究.通过一系列的实验验证了此次研究的真实性、可靠性和稳定性,从而为运动状态下的身份认证进行了有益的探索.
Based Authentication unfolding human characteristics still remain m the stationary phase ox cer- tification, using state of the human body segmentation and motion-based prediction methods expanded state of the body's identity. First, from the stationary state of the human body segmentation method is used to determine the basis of the skeleton; Secondly, the classical HMM theory to apply properly expanding mo- tion prediction of the human condition; Finally, with the above findings, the completion the motion state of identity authentication. Through a series of experiments in this study verified the authenticity, reliability and stability, so as to exercise the state's identity a useful exploration.
出处
《湘潭大学自然科学学报》
CAS
北大核心
2013年第2期80-85,共6页
Natural Science Journal of Xiangtan University
关键词
HMM
运动
身份
认证
人体
HMM
movement
identity
certification
human