台灣目前已邁入高齡化社會,老年人口預估在2040年將再增加至31%,進入「超高齡社會」階段。為確保老人的日常生活能力,體適能維持的問題將日趨重要。現階段的影像偵測及3D動畫技術軟體的成熟,有助於實現對於人體動作行為的捕捉以及復健動畫製作應用。本研究目的在於建立一套3D動作指引系統輔助衰弱老人日常生活動作行為訓練或復健期的體適運動,利用3D動畫的吸引力來提升增加患者做體適能活動的意願,鼓勵使用者持續進行體適能活動以維持身體機能。 本研究基於Kinect擷取動作者的關節點骨架資訊,再配合Blender所建立的3D人體模型合成出虛擬人物的動作,作為指引體適能動作的角色,而所得的資料亦可據以建立完整之動作資料庫供後續研究使用。目前研究結果完成兩位3D虛擬人物,使用者可控制3D人物及時反映其動作情形。使用者問卷調查結果顯示,本研究所開發之虛擬人物對吸引使用者進行體適能活動有正面的評價。未來研究可朝體適活動完成度判斷方面進行。
Taiwan currently has entered the aging society, older population estimates in 2040 will be further increased to 31% and enter "hyper-aged society". To ensure the activities of daily living of the elderly, maintain physical fitness will become increasingly important. With the mature of the image detection and 3D animation methods, it becomes possible to develop the techniques of human motion capture and rehabilitation animation. The purpose of this study is to establish a 3D movement guiding system to support fragile elderly assisted behavioral training activities of daily living or rehabilitation of physical fitness exercise. Attractive 3D animation is considered to promote patient willingness to do physical fitness activities. This study uses Kinect to track the human skeleton joints and then uses Blender to establish the required 3D human skeleton model. The human skeleton joints data can be used to build an action database for subsequent research. As a result, our research has built two 3D virtual characters which can be controlled by Kinect to immediately react to the action from the tester. User questionnaire survey result shows that the virtual characters developed in this research can attract users to conduct physical fitness activities. Future study may focus on the completeness assessment of physical fitness activities.