Performance improvement and height estimation of pedestrian dead-reckoning system using a low cost MEMS sensor

In this paper, PDR system's Improved performance and height estimation are introduced using a low-cost IMU. The personal navigation systems typically have used GPS. But the area where GPS signals are blocked should not be used such as indoor environments. To compensate for this, one of a technique is used PDR system. The location of a pedestrian estimates using only PDR system of the IMU. First of all, in order to estimate the location, a low-cost sensor corrected and manufactured the module. The step is detected using acceleration information by moving of the pedestrian through the module. A step length can be estimated using the detected step. The position estimation and heading will be estimated by the acceleration sensor and gyro sensor. Finally, information of the acceleration sensor and barometer applied to the KF(Kalman Filter) to estimate the height. This paper was tested in a real-environment based on the proposed algorithm. As the result, excellent performance was confirmed within the average distance error within 1% and the position error within 4% in the location estimate. And the results of height estimate, average standard error is confirmed improved performance as max 0.15m in case of using Kalman Filter.