| Summary: | 碩士 === 中原大學 === 電機工程研究所 === 100 === We use the Acceleration Sensor which is based on Zigbee in a Home-Care system, to detect hand gestures of the users. This can help the expression of senior citizens or patients to have a well conversation with care giver.
At first, as our system model, we apply an Acceleration Sensor which is based on Zigbee, and set it up to the back of the handedness of the care recipient to get the hand gesture. Then, in the operation of this system, when they want to talk about something, the system will be started and detect the accelerator by the acceleration sensor. According to the calculation, a locus will take shape in the 3-axis coordinate. In addition, we use our own method to add a concept of strength to pre-process this signal and use an algorithm, Longest Common Subsequence, to pair it with a structured learning system to sift a suitable word out. Finally, we display the result of the simulation that may suit for care recipient.
The contributions of our research are as follows:
(1) Realizable: It has low power consumption, can easily realize in the home environment.
(2) Practicable: This can help the senior citizens or the patients to speak to the care giver by themselves easily.
(3) Differentiable: Our method can distinguish the strength of the action.
(4) Expandable: To more delicate things, for example, sign languages, this thesis can give the data of the acceleration on the back of the hand.
|
|---|
