頁籤選單縮合
題 名 | Human Extremity Motion Analysis=人類肢體運動分析 |
---|---|
作 者 | 張意政; 黃仲陵; | 書刊名 | 影像與識別 |
卷 期 | 5:1 1999.03[民88.03] |
頁 次 | 頁20-30 |
分類號 | 312.1 |
關鍵詞 | 肢體運動分析; Body signature; Posture graph; Posture transition path; Model fitting process; |
語 文 | 英文(English) |
英文摘要 | This paper discusses the problems of the posture recognition in 2-D approach and posture identification in 3-D approach. With regarded to the posture recognition problem, a skeleton-based human walking motion analysis system is proposed to determine the posture type. The system consists of three major phases:(1) pre-processing phase, (2) model construction phase and (3) model analysis phase. In the first phase, it extracts the human body skeleton from the background and then obtains the body signatures. In the second phase, it analyzes the training sequences to generate the statistical models. In the third phase, it uses the trained models to recognize the input human motion sequence and calculate the motion parameters. For the posture identification problem, we propose a vision-based posture analysis system by using inverse kinematics method. Through obtaining the 3-D positions of five mark points (MPs) on the limbs, our system may accurately estimate the joint angels on the limbs. The approach is based on a 3-D data-model-fitting method that contains two model fitting processes:(1) the directional fitting process, which determines the moving direction, and (2) the extremity fitting process, which locates the angle parameters of the body joints and extremity joints. |
本系統中英文摘要資訊取自各篇刊載內容。