近年來智慧型手機使用量的大幅成長，重力感測器已經成為智慧型手機的標準配備，並且可以應用在體感動作偵測，提供使用者一個新的與介面溝通的方式。但是目前常見使用到重力感測器的系統，為了實做方便，通常限制了使用者持握手機的方向。 本論文提出一種在不限制使用者持握手機的條件下，利用手機上的重力感測器偵測人體的動作的方式。本論文並且應用拉邦動作分析來針對使用者的動作進行判斷與分類。 In recent years, due to the growth of the Smartphone users, the G-sensor has already become the standard equipment on Smartphone. Besides, G-sensor can be applied to the body motion sensing and provide new communicating interface for users. However, for easy implementation of gravity sensing system, the application would usually restrict users the way of holding Smartphone. In this thesis, we propose a method that can analyze the body motion without restricting the way users hold their phones by the G-sensor of the Smartphone. In this thesis, we analyze and classify the motions of users with LMA.
