Pedestrian Indoor Positioning Method Using Magnetic Data

Pedestrian Indoor Positioning Method Using Magnetic Data
复制标题

利用磁数据的行人室内定位方法

DOI:
--
复制
发表时间:
2016
期刊:
--
影响因子:
--
通讯作者:
H. Inamura
H. Inamura
中科院分区:
--
文献类型:
--
作者:
Yuya Murata;Katsuhiko Kaji;Kei Hiroi;Nobuo Kawaguchi;Takeshi Kamiyama;Ken Ohta;H. Inamura

文献摘要

被引文献

相似文献

在本研究中,我们提出了一种利用磁性数据进行室内定位的方法。在建筑物内观测到的磁数据往往会稳定多年,并与位置有关。该方法利用行人观察到的磁场数据,估计出行人的位置。使用行人磁场数据使数据收集变得更容易,并降低了位置估计成本。为了准确估计位置,我们建立了磁场数据模型。通过将路径划分为小单元来创建磁场数据模型。每个单元都有磁平均、磁标准差、连接和坐标。利用粒子滤波的方法,通过比较观测数据和磁场数据模型中的每一个单元来估计行人的位置。我们用HASC-IPSC和之前的实验结果对该方法进行了评估。结果表明,在特定条件下,该方法估计行人的位置误差在10m以内,与传统的指纹识别方法相比,数据采集成本降低幅度更大。
In this research, we propose an indoor positioning method using magnetic data. Magnetic data observed in the building tends to be stable for years and has location dependency. Proposed method utilizes magnetic data observed by pedestrians, and estimates pedestrian position. Using pedestrian magnetic data makes data collection easier and reduces position estimation cost. To estimate position correctly, we create magnetic data model. Magnetic data model is created by dividing routes into small cells. Each cell has magnetic average, magnetic standard deviation, connection and coordinates. Using particle filter, we estimate pedestrian position by comparing observed data and each cell of magnetic data model. We evaluate this method with HASC-IPSC and the result of prior experiment. The result shows that proposed method estimates pedestrian position with error within 10m under a specific condition, and confirms that the data collection cost is reduced larger than conventional Fingerprinting method.