隨著軟硬體效能的快速演進，動作捕捉系統近年來又成為關注的課題，系統將人體 的動作轉換成數位資料，並進行後續的分析與應用，除了可以簡化虛擬場景建置動態人 物的資訊處理外，並常見於優秀運動員的動態分析；近年來動作捕捉系統可以用更經濟 的軟硬體建置，除了促使動作捕捉的技術更成熟與普及外，各種創新應用也持續被開發 出來。 本研究運用微軟公司在 2010 年底推出的Kinect，以紅外線3D 感測技術與對應的軟 體模組，建置偵測骨架追蹤系統，運用其技術開放與經濟的特性，開發學童骨架動態資 料蒐集系統與初步分析模組，透過較為精細動作偵測，提供體適能、學習情緒等可能影 響活動特徵進一步分析比較的參考，初步結果顯示：配合系統說明可依據受測者身長與 運動目標架設最適化的系統架構，提供較完整的資料蒐集，典型的電腦可同時提供多台 Kinect 進行更大範圍的監測；進入受測區後之系統反應時間約0.4 秒，顯示多台Kinect 可有改善進測範圍，此外平均速度的穩定性，可以用來評估受測人的特質甚至可能與受 測者心情有關連，而本系統提供各肢體的運動軌跡，與動態分析，預定後續進行學童體 適能的關聯性分析。

Recently researches pay attention to motion capture systems since the performance increased in both of hardware and software. The system digitalizes human action and then further analysis and application may carry on by computers. The economical way for motion capture stimulate the progress of its hardware and software technologies and then more creative applications are expected. This study employed Microsoft integration development environment and toolkit to build a skeleton tracing system including the detection of skeleton in motion, the in-situ storage module and some of analysis and presentation modules. The object of this system are designed for teachers in the primary school to collect students motion characters and use to adaptive instruction The preliminary testing was carried by an adult and a primary school student while the results show the significant different in their walking behavior and the detection lag of skeleton stream is about 0.4 second that may be various due to the hardware performance.