體感姿態辨識應用於營造業姿態性累積職業傷害評估之研究

Abstract

累積性職業傷害不同於其它職業災害(墜落、感電等)，其危害並非立即顯現，多數為骨骼肌肉疾病，且初期之痠痛疲勞症狀容易被忽略，然而經年累月後卻可能對勞工產生永久性的失能傷害。造成累積性傷害的主要原因多為不良或錯誤的作業姿勢，若能及時性的評估姿勢的危害等級且讓勞工自我了解危害，則可大幅降低累積性職業傷害的發生率。 有鑒於近年來影像感應技術的進步，以攝影方式即可進行動態追蹤，其中由Microsoft所推出的Kinect體感器，可藉由拍攝人體之動作進行操作且不依靠任何儀器並追蹤其骨架，並將其擴展至人因作業領域，應用於累積性職業危害評估之中。 整合體感姿態辨識及人因作業姿勢分析技術，研發營造作業姿勢自動化評估系統。本論文將針對營造業較常發生累積性職業傷病之姿勢動作進行系統驗證及修正，並評估系統在教育訓練及醫療協助之適用性。

The damage of cumulative trauma disorders is not as apparent as other occupational accidents such as falling, and can be easily ignored in the construction industry. However, the accumulation of the damage over the time may cause permanent disabling injuries. Most cumulative trauma disorders are caused by inappropriate postures, and the related injuries may be effectively reduced if workers correct their inappropriate working postures. With the rapid development of depth image sensor technology, we can using Kinect to do dynamic tracking posture , which is develop by Microsoft.It can track the skeleton and determine the position of the joint without wear any sensors. To extend it to human factors, applied to the damage of cumulative trauma disorders This research integrates motion sensing and capture techniques, and ergonomics work posture identification with multimedia training materials as a whole system. The research will also perform an experiment involving the working postures highly relevant to cumulative trauma disorders (e.g., carrying, lifting, and stooping). The evaluation of the proposed system in training, diagnosis, and rehabilitation will also be conducted