利用智慧型手機進行頸部姿勢之監控

Abstract

近年來智慧型手機對大眾的影響甚鉅，不僅手機持有率上大幅上升，民眾使用智慧型手機的時間也大幅增加，日常生活中「低頭族」隨處可見。然而多數人使用智慧型手機的不良姿勢容易導致肩、頸和背部相關的痠痛與疾病。因此本研究利用手機前鏡頭開發一個監測使用者頭部角度的系統藉以達到提醒使用者不要長時間維持於同一個姿勢之目的。 本篇研究設計了監測系統，每隔一段時間系統會根據手機前相機拍到的使用者臉部進行角度估計，綜合當下手機角度去監測在時間區間中最大與最小角度相減是否有超過門檻值，如有我們定義使用者在時間區間內有改變姿勢，如沒有則視為時間區間內皆沒有變換姿勢，則使用者會收到系統端給予的提醒訊息。本篇研究共設計四個實驗，第一個實驗我們設計兩項姿勢並驗證頭部角度估計的準確性，並探討兩個姿勢中使用者的角度變化。我們進行另外兩個實驗來設定監測系統的參數。最後一個實驗則是找了三名使用者使用系統並確實監測他們的姿勢。

In recent decade, smartphone has a great influence on human’s life. Smartphone ownership grows at a skyrocketing speed, and people spending more and more time on using smartphone every day. But the poor posture and overusing of smartphone can easily cause neck pain. There’s also a new term called ‘text neck’, which means bending neck forward for a long period of time or too frequently. There’s several study shows that neck flexion leads to neck pain. In this paper, we propose a prototype monitoring system to remind people not to stay in a static posture too long. The monitoring system capture users’ face from smartphone’s front camera and check whether participants are stay in a static posture or not. In a period of time, we use the maximum head tilt angle to miner the minimum head tilt angle, if it is over the threshold angle, then we think participants have moved; otherwise, the server would send a warning message to remind users. We design four experiments in the paper. In the first experiment, there are two posture types which are the commonly two types of posture that people would be when they use smartphone. We investigate the difference of participants’ head tilt angle and also validate the accuracy of face orientation estimation method in experiment one. We set the parameters of the monitoring system by the another two experiments. In the last experiment, we found three participants to use our system for an hour while the system really reminded them when they stay in a static posture.