Design of an Ontology-Lite-Based Pre-Diagnosis Model and Significant Rare Data for the Six Phases of Golf Swing Injury Detection and Evaluation

View Statistics Email Alert RSS Feed

Information

Details

Project title

Code/計畫編號

MOST106-2221-E019-075

Translated Name/計畫中文名

利用個人化重要稀少性因子建構輕量化本體論預診機制應用於高爾夫六階段揮桿運動傷害模式

Project Coordinator/計畫主持人

Hao-Hsiang Ku

Funding Organization/主管機關

National Science and Technology Council

Co-Investigator(s)/共同執行人

歐俠宏

Department/Unit

Department of Computer Science and Information Engineering, Hwa Hsia University of Technology

Website

https://www.grb.gov.tw/search/planDetail?id=12487756

Year

2017

Start date/計畫起

01-08-2017

Expected Completion/計畫迄

31-07-2018

Bugetid/研究經費

530千元

ResearchField/研究領域

資訊科學--軟體

高爾夫球運動已逐漸普及化、年輕化與全民化，是一項非常注重身體協調性、平衡性以及標準性的運動，強調揮桿時身體的姿勢，不標準的揮桿動作易造成運動傷害，因此預防高爾夫球錯誤揮桿姿勢造成運動傷害儼然已成為被關注的重要議題之一。高爾夫球揮桿動作可分為預備期、上桿期、下桿期、加速期、送桿期與收桿期六個階段，六個階段均易有姿勢不正確的問題產生。有鑑於此，本計畫規畫與設計一輕量化本體論預診機制應用於高爾夫六階段揮桿運動傷害模式(Ontology-lite-based Pre-diagnosis Model and Significant Rare Data for the Six Phases of Golf Swing Injury Detection and Evaluation, OPM4Golf)透過動作預診提早發現錯誤姿勢，預防揮桿運動傷害發生。OPM4Golf 為三層式架構環境，包括(1) 具輕量化推論機制之行動式使用者端(Golfer)，(2) 具六階段高爾夫球揮桿姿勢分析模型之動作觀測應用伺服器(Motion Observation Server)以及(3) 具重要稀少資料分析之大數據推論模式雲端伺服器(Big Data Server)，規劃出6 種揮桿模式與18 種錯誤揮桿傷害姿勢，透過整合輕量化本體論(Ontology-lite) 技術加快建模與運算速度，建構起個製化(Personalized) 揮桿動作預檢 (Pre-Diagnosis)機制。位於雲端之大數據推論模式雲端伺服器以Hadoop 大數據資料庫建構建構起大眾行為資料，再以本研究所提出之增權式不平衡學習法(Weight-added Class-imbalance Learning)進行稀少資料分析推論，整合個別輕量化本體論模式滿足不同使用者之情境、特性與需求，建構起預診模式 (Pre-diagnosis)。高爾夫球員透過整合Kinect 體感感測器及穿戴式三軸加速規感應設備蒐集揮桿過程中的揮桿動作，再將資料送至動作觀測伺服器中運用所提出之高爾夫球揮桿姿勢分析模型(Golf Swing Posture Analysis, Golf-SPA)進行分析揮桿數據，檢測出可能發生運動傷害之不正確的姿勢動作，並發出警示通知運動員，以達成預防運動傷害的目的。有鑑於上述內容，本計畫主要為規劃與建置高爾夫球員揮桿行為模型、行動設備上之輕量本體論(Ontology-Lite)、六階段高爾夫錯誤姿勢與傷害影響模式、輕量化本體論揮桿預診模式、重要稀少因子選定機制、增權式不平衡學習法等，並將其模式進行模組化工作。最後，本計畫所產出之 OMP4Golf 除可當作一未來研究的參考模型外，亦將移植到國立體育大學測試使用。 Golf is a kind of high rhythm exercise, which emphasis on the body balance, coordination and standard. Golfers need to concentrate on swing postures. Non-standard swing is easy to come about sports injuries. Hence, this project designs and proposes an Ontology-lite-based Pre-diagnosis Model and Significant Rare Data for the Six Phases of Golf Swing Injury Detection and Evaluation, which is called OPM4Golf. A swing posture is divided into six phases, which are preparation, takeaway, forward swing, acceleration, early follow-through and late follow-through. The reference engine of OPM4Golf is based on ontology, artificial neural network and Weight-added Class-imbalance Learning (WCL) method which constructs mass behaviors model and ontology-lite for Golfers. A Golfer is with a smartphone and Kinect sensors to detect the Golfer’ swing actions and to interact with OPM4Golf. A Motion Observation Server is with mass behavior model and Golf Swing Posture Analysis (Golf-SPA) model to check Golfers’ swing actions and to alter Golfers when they are with error actions. Big Data Server is based on the Hadoop database, which uses MapReduce and Weight-added Class-imbalance Learning to manipulate Golfers’ raw data and influence factors and to construct pre-diagnosis model for this Golfer. Hence, this project will define and design the mass behavior model, the ontology-lite model, the Golf Swing Posture Analysis (Golf-SPA) model, the pre-diagnosis model and the six phases of Golf swing injury detection and evaluation of OPM4Golf. Finally, OPM4Folf can be a reference model for researchers and engineers.

Keyword(s)

預診模式
輕量化本體論
大數據分析
高爾夫揮桿
運動傷害
Pre-diagnosis
Ontology-lite
Big Data Analysis
Golf Swing
Sport Injury

DSpace CRIS

Design of an Ontology-Lite-Based Pre-Diagnosis Model and Significant Rare Data for the Six Phases of Golf Swing Injury Detection and Evaluation

Details

Description