Design of Mobile Application Control Based On Hand Gesture

In recent years, we have seen dramatic progress in the field of Hand Gesture Recognition, yet these applications have some drawback. Keeping in view the trend of making application more interactive and simple, easy to use as well as cost efficient, a Hand Gesture Recognition based system is developed to allow interactive control of application using hand gesture. This paper presents a framework for hand gesture recognition based system which controls the mobile application. The working of this framework is based on the information fusion of 3 Axis Accelerometer (ACC) and Surface Electromyography (SEMG) sensors. A Gaussian Mixture Model (GMM) and Fast Dynamic Time Warping Algorithm (FDTW) are utilised to get the final result. In addition the system including a wearable hand gesture sensing device which is worn on the forearm and an algorithmic framework for the smart phone to recognise the hand gesture. The Armband which is worn on the forearm is able to manipulate 11 predefined gesture or even the personalised ones. The results suggest that the developed prototype responded to each gesture within 300 ms on the mobile phone, with average accuracy of 96%.