透過您的圖書館登入
IP:18.220.1.239
  • 期刊

Motion Analysis of Facial Animation by Video-Based System-A Preliminary Result

利用攝影系統來做顏面表情運動的動態分析-初步報告

摘要


利用運動學原理來定量評估上肢及下肢的正常活動與病態運動,在近十幾年來已被廣泛的運用於外科、骨科、復健科、醫學工程以及運動科學領域,然而對於顏面表情動作的運動學研究仍十分缺乏。我們嚐試利用攝影系統來分析正常人的五種表情動作,如微笑、抬眉、閉眼、鼓腮以及吹口哨。結果顯示一般利用照相機或重複曝光攝影機等二維空間計量方式常低估表情運動範圍,尤其是顏面反光球的運動如果是走在前後平面時最容易發生,其誤差百分比可以達到五十七,因此三維空間的定量分析才是正確而可行的方法,用以評估正常人的顏面表情運動及顏面神經麻痺患者的治療前後定量分析。

關鍵字

無資料

並列摘要


Three dimensional motion analysis was widely used in the kinematic study of upper and lower extremities disorders, such as cerebral palsy, stroke patient…. But application of this method to the study of facial animations was rarely utilized. We tried to use the video-based system (VICON system) to evaluate the quantitative analysis of five facial move-ment: eyebrow lift, eye closure, smile, puffy face and whistle. The preliminary result revealed that 2D amplitudes significantly underestimated the 3D amplitudes for all animations. When the 2D amplitudes were expressed in the percentage of the 3D amplitudes, however, the difference were as large as 57% for the left lateral canthus point during eye closure. So we recommended that 3D analysis by video-based system is much more appropriate for detecting the differences in facial movement than 2D analysis by the image of camera.

延伸閱讀