针对手机加速度传感器采集的数据特点,结合小波变换理论,提出一种基于小波多层分解的提取手势轨迹特征并用于身份认证的方法。采集手机加速度的原始数据并预处理,利用小波多层分解得到每一层高频信号和最后一层低频信号的能量组合为特征向量,用S V M分类器进行鉴别模型训练,进而实现手机端的身份认证。研究结果表明:所提出的基于手机加速度传感器的身份认证方法准确率达9 6%以上,具有较好的推广价值。
According to the features of the data collected by mobile phone accelerometer sensor,combined with the wavelet transform theory,a method based on multi-layer decomposition of wavelet layers was proposed to obtain gesture trajectory feature for identity authentication.By collecting th eoriginal data,the every layer of high frequency was obtained from wavelet decomp osition and it was combined with the last low frequency signalt obuild the feature vector,and then the identification model was trained using SVMcl assifier,and its authentication was achieve don mobile phone.The results show that the identity authentication method has the accu rate rate of 96% and has good promotional value.