- 期刊
利用G-Sensor及電子羅盤於手機定位之實驗
Using G-Sensor & Digital Compass for Device Positioning-An Experiment on Google Android Dev Phone
摘要
近來行動裝置上之感應裝置越見多樣化，如加速器、氣壓計、電子羅盤等不一而足。於室內使用行動裝置進行導航，則室內必須建置一定AP數量以上的無線網路；若加上慣性導航系統來加以輔助，便能夠在原本定位系統之上提供校正資訊、或系統失效時提供暫時的定位結果；反之在一般無法建置智慧型環境的情境下使用慣性導航，則成為較為合理的室內定位解決方案。許多研究嘗試整合高精準度之感應器來偵測使用者動作，但其成本過高，並不能跨入尋常百姓家讓一般民眾使用。本文將以實際出現在消費型裝置上的感應裝置：Google Android Dev Phone 1作為實驗機器，讀取該機器上包含三軸加速器、電子羅盤、磁力計、GPS、Wi-Fi五種訊號來進行手機定位之實驗。 本研究於台北共六個區域進行步行實驗。研究發現ADP之電子羅盤數據會對最終路徑計算結果造成極大影響，微小的角度變化在持續行走之後都會造成路徑嚴重的偏移，但以三軸加速度所繪製出來的行走路徑與GPS訊號紀錄的路徑為相似形，且計算成本不高未來可實際應用於移動裝置上。行走模式的判斷會是為來研究的主要目標，尤其在裝置無法隨時固定在同一位置時更顯得該研究方向的重要性。
並列摘要
Applying inertial navigation system to indoor navigation are considered a supplement way of how to improving indoor positioning in a smart place. Many researches try to detect human motions by integrating multiple high sophisticated sensors into one device. On the other hand, sensors of lower price have been integrated within mobile devices, like Google Android Dev Phone. In this study, Google ADP will be used as experiment device, to collect accelerometer, digital compass, magnetic field, GPS and Wi-Fi signals along those routes we choose in certain areas of Taipei City. Many modifications are needed for tuning how to distinguish the walking patterns of human because of somehow not-so-precise sensors on ADP. In this study, we find that digital compass could be the main factor that affects the path drawn by our path calculation method. Minor angle changes of digital compass could lead to completely different direction until next turn taken. However the total length of walking path are measured quite similar with the path drawn by GPS signals, which we could use in daily life without intensive computing efforts.