本論文提出一個以慣性感測器為基礎之可攜式裝置及其行人導航演算法來加以估測行人在室內環境中的行走軌跡。此可攜式裝置是由一個三軸加速度計、一個三軸陀螺儀、一個三軸磁力計、一個微控制器及一個藍芽無線傳輸模組所構成。使用者在室內環境配戴此裝置以正常速度行走進行行人導航。此行人導航演算法包含了步數估測、方位角估測、步長估測及行走軌跡估測等四個程序，其主要是用來處理所量測的慣性感測訊號及估測行人在室內環境中之行走軌跡。此外，為了降低慣性感測器的誤差累積，我們利用互補式濾波器進行姿態誤差補償，來加以降低慣性感測器本身內部所產生的誤差。本文所開發之行人導航系統包含了以下優點：(1)此系統可以在室內環境中不需任何室內定位技術輔助下進行行人導航及(2)此行人導航演算法可有效地降低慣性感測器的積分誤差並可以準確地估測行走軌跡。最後，經由室內環境行走軌跡估測的實驗結果，終點誤差改善了50.4%~73.8%，成功地驗證此可攜式裝置及其行人導航演算法之有效性。

This paper presents an inertial-sensor-based wearable device and its associated pedestrian navigation algorithm to estimate pedestrian trajectories in indoor environments. The wearable device is composed of a
triaxial accelerometer, a triaxial gyroscope, a triaxial magnetometer, a microcontroller, and a Bluetooth wireless transmission module. Users can carry the device to walk in indoor environments at normal speed for
pedestrian navigation. A pedestrian navigation algorithm composed of the procedures of stride detection, heading angle estimation, stride length estimation, and pedestrian trajectory estimation has been developed for estimating pedestrian trajectories in indoor environments. In order to minimize the integral errors caused by the intrinsic noise/drift of inertial sensors, we utilized the complementary filter for correcting orientation estimation. The advantages of this pedestrian navigation system include the followings: 1) It can be used in indoor environments without any external indoor localization technique; and 2) the pedestrian navigation algorithm can reduce integral errors effectively and thus can estimate pedestrian trajectories accurately. The accuracy can be modified 50.4%~73.8% of the finishing point on pedestrian trajectory estimation. This experimental result has validated the effectiveness of the inertial-sensor-based wearable device and its pedestrian navigation algorithm.