手机加速度计的行人行进状态识别

Pedestrian movement status identification based on mobile phone attitude sensor

针对行人在行进过程中会出现后退而导致对行人航位轨迹的错误判断问题,该文以手机内加速度传感器信号为数据依据,以识别行人正常前进中的后退状态为研究目标,研究三轴信号的均值、方差、轴间协方差等时域特征,采用经验模态分解、最大相关最小冗余、最小二乘支持向量机等方法,识别行人实时的前进或后退。研究结果表明:两人将手机置于不同位置,分别采集前进中出现不同后退步数的实验数据,以一组为建模数据,识别其他情况的运动状态,其识别平均成功率达96.00%,具有较大的理论参考价值。

Aiming at the problem that pedestrianvS would retreat during the process of travel,resulting in incorrect judgment of the trajectory of pedestrians,empirical mode decomposition,max-relevance and min-redundancy,and least square support vector machine were used to identify pedestrians9 forward or backward in real time in this paper,by taking the accelerometer signal in the mobile phone as the data basis,and recognizing the pedestrian^s retreating state during normal forwarding as the research goal,and studying the time-domain features such as the mean,variance,and covariance of the three-axis signals.The research results showed that:two carriers with mobile phone being put in different body parts respectively to collect the experimental data when there were different backward steps in forward walking.And according to one group of data,it set a modeling data to identify the movement state in other condition,and the recognition success rate reached to 96.00%9 with a larger theoretical reference value.