即時摔倒辨識回報警示服務系統

Real-time fall recognition warning service system

邱奕晨 ； 龍祈澔

《數位媒體設計國際研討會論文集》 第十四屆 (2021/06) Pp. 333-336

摘要

每年因為摔倒受傷甚至是死亡的人數非常多，經過衛生福利部統計於2019年各類事故傷害死亡人數，跌倒死亡僅人數僅次於交通事故位居第二，如何防範以及事故發生後的處理成為現在各個公司與政府關注的議題。本研究分為三大部分，首先是影像辨識以及LineBot連動，在此會講解系統運算的發法的以及資料傳輸的運作。接下來是服務設計流程，本部分探討回報流程中的各個時間點以及觸發角色之間關係，以及各角色如何與系統互動。最終是後續資料延伸，透過分析紀錄資料，進行事故預防與危險因子移除動作。實驗針對市內居家老人跌倒防範進行事件模擬與測試，紀錄各合作人訊息反應時間以及回報時間點確認該系統是否能提升事故發生效率，透過數據以及影像的紀錄，我們可以了解事故發生的原因並進行事故再發生可能行排除的動作。

關鍵字

OpenCV ； 影像辨識 ； Webhook ； LineBot ； IOT

並列摘要

There are a lot of people injured or even killed by falls each year. According to the Ministry of Health and Welfare statistics, the number of deaths from various accidents in 2019. The number of deaths from falls is second only to traffic accidents. How to prevent and deal with accidents It has become a topic of concern for companies and governments. This research is divided into three parts. The first is image recognition and LineBot linkage. Here we will explain how the system calculates and how to transmit data. Next is the service design process. This section discusses the relationship between each point in the information transmission process and the triggering role, and how each role interacts with the system. Finally, it is the follow-up data extension, through the analysis of the record data, the accident prevention, and the removal of dangerous factors. The experiment conducted event simulations and tests on fall prevention for the elderly at home in the city and recorded the response time of each partner's message and the reporting time point to confirm whether the system can improve the efficiency of the accident. Through the data and image records, we can understand the cause of the accident and Perform actions that may eliminate the possibility of recurring accidents.

並列關鍵字

OpenCV ； Image recognition ； Webhook ； LineBot ； IOT