Abstract: In order to guarantee the security of sensitive information in intelligent mobile phone, a 3D gesture authentication scheme based on a mobile phone with three-axial accelerometer embedded is put forward. On the basis of analyzing the energy distribution characteristics of gesture acceleration signal, a new gesture endpoint detection algorithm based on energy entropy was proposed. Euclidean Dynamic Time Warping（EDTW） method is used to analyze these temporal signals after gesture endpoint detection for authentication, False Rejection Rate（FRR）is around 7 percent while the False Acceptance Rate（FAR） is approaching 0 percent, so intelligent mobile phone user identification can be achieved.

Key words: 3D gesture authentication, gesture endpoint detection, Short Time Fourier Transform（STFT）, energy entropy, Dynamic Time Warping（DTW）

摘要： 为了保证智能手机敏感信息的安全性，设计实现了一种基于手机内置三轴加速度传感器的三维手势认证方案。在手势端点检测部分，在定性分析手势加速度信号能量分布特性的基础上，提出了一种基于能量熵的新方法实现有效手势截取。进一步设计基于欧式距离的动态时间规整算法对截取后的手势序列信号进行匹配认证，当他人模仿手势错误接受率趋近0%时，本人认证手势错误拒绝率维持在7%左右，从而实现智能手机用户身份识别。

关键词: 三维手势认证, 手势端点检测, 短时傅里叶变换, 能量熵, 动态时间规整