結合智慧手機與Kinect之跌倒偵測監控平台開發

Developing fall detection reporting system by an integration of smart phone and Kinect

台灣慢慢已存在高齡化社會的現象，所以老年人口的照顧是一個值得研究的問題，而跌倒是老年人口普遍存在的問題，針對此問題，本研究考量以行動導向服務（Location-based services）進行老年人跌倒偵測與即時通報，滿足老年人照顧的需求。本研究主要著重於跌倒偵測監控平台之開發，以結合智慧型手機及Kinect等技術，由智慧型手機作為行動端跌倒偵測，考量手機內的三軸加速計提供的資訊即時偵測老人跌倒，也以定點裝置室內Kinect來偵測跌倒行為，透過Kinect偵測的骨架節點資訊，得到頭部速度與高度，來判斷是否跌倒，最後整合行動端與Kinect兩者跌倒偵測資訊確認是否真正跌倒，並可回傳老人跌倒影像、時間與空間位置至WebGIS跌倒偵測監控平台。 Since the ageing society exists in Taiwan, the elderly population care is worth noticing. Obviously, falls in the elderly is a common problem. To deal with the problem, this study considers location-based service for fall detection and reporting of the aged. This study integrated smart phone and Kinect to fall detection, and provided fall down information for senior care. This study applied a smart phone as the mobile platform and applied accelerometers to detect the fall down in the aged. Moreover, the Kinect was applied to detect the fall down in the room based on head velocity and height. Fixed and location-based sensors can be combined for obtaining the sufficient fall down information. Based on the images, location and time of fall detection, the results were demonstrated in the Web GIS, and were sent to relatives and care units.

Since the ageing society exists in Taiwan, the elderly population care is worth noticing. Obviously, falls in the elderly is a common problem. To deal with the problem, this study considers location-based service for fall detection and reporting of the aged. This study integrated smart phone and Kinect to fall detection, and provided fall down information for senior care. This study applied a smart phone as the mobile platform and applied accelerometers to detect the fall down in the aged. Moreover, the Kinect was applied to detect the fall down in the room based on head velocity and height. Fixed and location-based sensors can be combined for obtaining the sufficient fall down information. Based on the images, location and time of fall detection, the results were demonstrated in the Web GIS, and were sent to relatives and care units.