Carrying Position Independent User Heading Estimation for Indoor Pedestrian Navigation with Smartphones.
Carrying Position Independent User Heading Estimation for Indoor Pedestrian Navigation with Smartphones.
复制标题
使用智能手机进行室内行人导航的与携带位置无关的用户航向估计
DOI：
10.3390/s16050677
复制
发表时间：
2016-05-11
期刊：
影响因子：
--
通讯作者：
Cui Y
中科院分区：
文献类型：
--
作者：
Deng ZA;Wang G;Hu Y;Cui Y
提出了一种利用智能手机内置惯性传感器进行室内行人导航的航向估计方法。与以前的方法约束携带位置的智能手机在用户的身体上，我们的方法给用户一个更大的自由度，通过实现自动识别的设备携带位置和随后的选择的最佳策略的航向估计。我们首先使用加速度计和气压计的决策树预先确定的运动状态。然后，为了实现准确和计算量小的携带位置识别，我们将位置分类器与新颖的位置转换检测算法结合联合收割机，该位置转换检测算法也可以用于避免在行人行走期间位置转换和用户转弯之间的混淆。对于放置在裤子口袋中或握在摆动的手中的设备，通过部署基于主分量分析（PCA）的方法来实现航向估计。对于在电话呼叫期间握在手中或抵靠耳朵的设备，通过将设备的偏航角添加到相关航向偏移来直接估计用户航向。实验结果表明，该方法能够自动检测出载体位置，具有较高的准确性，在准确性和适用性方面优于以往的航向估计方法。
This paper proposes a novel heading estimation approach for indoor pedestrian navigation using the built-in inertial sensors on a smartphone. Unlike previous approaches constraining the carrying position of a smartphone on the user’s body, our approach gives the user a larger freedom by implementing automatic recognition of the device carrying position and subsequent selection of an optimal strategy for heading estimation. We firstly predetermine the motion state by a decision tree using an accelerometer and a barometer. Then, to enable accurate and computational lightweight carrying position recognition, we combine a position classifier with a novel position transition detection algorithm, which may also be used to avoid the confusion between position transition and user turn during pedestrian walking. For a device placed in the trouser pockets or held in a swinging hand, the heading estimation is achieved by deploying a principal component analysis (PCA)-based approach. For a device held in the hand or against the ear during a phone call, user heading is directly estimated by adding the yaw angle of the device to the related heading offset. Experimental results show that our approach can automatically detect carrying positions with high accuracy, and outperforms previous heading estimation approaches in terms of accuracy and applicability.