Application of 3D Accelerometer for Automatic Recognition System of Physical Activities

• Main Authors: Chan shin-an, 陳信安
• Other Authors: 魏兆煌
• Format: Others
• Language: zh-TW
• Published: 102
• Online Access: http://ndltd.ncl.edu.tw/handle/12567278955219358112

碩士 === 南台科技大學 === 電機工程系 === 101 === Human motion analysis and recognition are important research topics of the modern kinematics and the medical science, such as the head or eye movements, opening or closing the lips, arms swinging, fingers or body bent, walking posture and balance and so on. According to these movement information of the body parts it can be used to develop mobility aids for disabled people, for assessment of patient rehabilitation, and analysis of exercise effectiveness. Person's movement is mainly achieved by legs, and many sports are completed through moving of the legs also, so the movement of the legs related to human's mobility, and the walking ability is also related to individual's health. In individual action, the pedometer is a simple equipment to calculate personal walk steps and inspire people to challenge themselves, enhance physical fitness, help body slimming. The early pedometer design uses a weighted mechanical switch to detect steps together with a simple counter, When it is shaken one can hear a metal ball sliding back and forth, or use a pendulum striking stops as it swings, and one strike for one count. This kind of pedometers are affected easily by a slight vibration, resulting to produce unnecessary errors. New art pedometer uses 3D accelerometer to sense the acceleration of the device, the number of steps can be determined after calculation and quantization of the sensing data. This pedometer can not only reduce the error, but also reduced the volume of the device significantly, make it easy to carry, but also open to develop other applications, therefore, this paper uses 3D accelerometer of the new pedometer, and mount it on the users ankle to retrieve instant acceleration data, then the data will be transferred to a computer via Bluetooth interface for recording and analysis, and then the user's bodily motion gesture can be determined, the effeteness of exercise can more accurately assess also, if more other physiological sensors are involved.