基於加速度特徵值之模糊手勢識別系統

智慧型行動裝置的普及與其功能日新月異，使得智慧型行動裝置已成為人們日常生活中不可或缺的產品。智慧型行動裝置內嵌各式感測元件，例如：加速規、陀螺儀或光感測器等，其中以加速規最為普遍。而傳統的手勢識別系統須依靠特製的裝置才能運作。近年來以智慧型行動裝置為主的手勢動作識別相關研究已提出來，但其仍有識別準確度不高及系統複雜度高的缺點。因此，本研究將提出一個基於手勢動作加速度特徵值之手勢識別演算法，並結合模糊推論的方法以提供高識別準確度及低系統複雜度之手勢動作識別系統。實驗模擬結果顯示所提之方法在識別向左傾斜、翻轉、左右搖擺和Z 字型等四種動作具有高識別準確度。

Because of the popularity of smart mobile devices and their functions change rapidly, smart mobile devices have become an integral part of people's daily life products. Smart mobile devices will be embedded in all kinds of sensing elements, for example: accelerometers, gyroscopes or optical sensors, of which the accelerometer is most common. Traditional gesture recognition systems have to rely on a special device to operate. In recent years, smart mobile devices based gesture recognition research has been proposed, but it had low recognition accuracy and high system complexity. Therefore, this paper proposed a acceleration characteristics based gesture recognition algorithm, and it also combined with fuzzy inference method to provide high recognition accuracy and low system complexity. Experimental results show that the proposed method in identifying four gestures such as tilted to the left, flip, swing around and Z-shaped has high recognition accuracy.