開發慣性感測器為基礎之動作追蹤方法與虛擬實境健身房系統建立;Development of IMU-based motion tracking method for exercise posture analysis in a virtual reality fitness environment

NCUIR > College of Electrical Engineering & Computer Science > Graduate Institute of Electrical Engineering > Electronic Thesis & Dissertation > Item 987654321/90069

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/90069

Title:	開發慣性感測器為基礎之動作追蹤方法與虛擬實境健身房系統建立;Development of IMU-based motion tracking method for exercise posture analysis in a virtual reality fitness environment
Authors:	黃致傑;HUANG, JHIH-JIE
Contributors:	電機工程學系
Keywords:	慣性感測單元;虛擬實境;健身房;逆向運動學;DTW;IMU;virtual reality;Gym;inverse kinematics;DTW
Date:	2022-08-26
Issue Date:	2022-10-04 12:09:56 (UTC+8)
Publisher:	國立中央大學
Abstract:	在疫情時代下，元宇宙蓬勃發展和遠距學習成為主流工具，本研究開發了一套雲端健身系統，結合穿戴式慣性姿態感測器（IMU）及VR頭戴顯示器，以Unity為虛擬平台，在虛擬世界中進行即時運動骨架分析，以實境運動場館VR虛擬健身教練的呈現方式進行健身動作示範與運動姿勢校正之建議，修正使用者運動姿勢以提升運動效率，並加以記錄和分析使用者運動數據。本系統使用12顆慣性感測器，將其配戴在受試者的四肢、腰部及背部，將四元數旋轉數據傳輸致電腦後，以逆向運動學方法計算，準確的獲取人體骨骼運動姿態，並以3D虛擬人物呈現，並可將運動數據紀錄做後續分析。在使用者畫面上，除了教練與自身的3D虛擬骨架外，更實時呈現各部位角度數值及姿勢校正的文字提示。在運動數據分析上，本研究將各部位的運動軌跡用動態時間規整方法(DTW)，將使用者與教練標準動作做相似度分析。本研究的成果經實驗證實，五位受試者經使用本系統後，從原本的錯誤動作，逐漸與標準動作相近甚至貼和，並且避免錯誤的姿態造成運動傷害。這系統除了可以應用在健身房，更可以進一步延伸應用到復健中心，甚至是長照中心，讓使用者在各所需的場景下進行正確的活動，獲得有益身心的成果，結合科技和運動進而實現全民遠端自主運動。;In the era of epidemic, the metaverse is flourishing and distance learning has become a mainstream tool. This research developed a cloud fitness system, combined with wearable inertial attitude sensor (IMU) and VR head-mounted display, using Unity as a virtual platform, and perform real-time motion skeleton analysis in the virtual world, demonstrate fitness movements and suggest exercise posture correction in the form of VR virtual fitness coaches in real sports venues, correct the user′s exercise posture to improve exercise efficiency, and record and analyze the user sports data. This system uses 12 inertial sensors, which are worn on the limbs, waist and back of the subject, and after the quaternion rotation data is transmitted to the computer, it is calculated by the inverse kinematics method to accurately obtain the motion posture of the human body, and presented as a 3D virtual character. The movement data can be recorded for subsequent analysis. On the user′s screen, in addition to the coach and its own 3D virtual skeleton, the text prompts for the angle value and posture correction of each part are displayed in real time. In the analysis of sports data, this study uses the dynamic time warping (DTW) method for the motion trajectories of each part to analyze the similarity between the user and the coach′s standard movements. The effectiveness of this study have been confirmed by experiments. After using the system, the five subjects gradually moved closer to the standard movements from the original wrong movements, and avoided sports injuries caused by wrong postures. In addition to being used in gyms, this system can be further extended to rehabilitation centers and even long-term care centers, allowing users to perform correct activities in each required scenario and obtain beneficial physical and mental results. Combining technology and sports to realize remote autonomous sports for all.
Appears in Collections:	[Graduate Institute of Electrical Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	35	View/Open

社群 sharing

Loading...