Virtual Reality(VR) is a popular segment that is rapidly developing in the recent years with the advancements in compact and cost-effective GPU technologies required to render such intensive tasks in real time. Communication in VR is usually done either through Voice or Text through a point and click method using motion controllers. It’s tedious to move and point towards a specific character to type a particular word and also breaks the illusion of the VR. In this paper we have proposed a method to counter these issues, by using gesture recognition of hand movements, using accelerometer and gyroscope data, through Machine Learning. This system easily integrates into the existing platform of motion controllers as they already include all the required sensors built into them already. A prototype is also showcased using an Arduino controller and an MP 6050 (Accelerometer and Gyroscope sensor) with data transmission through Bluetooth. This paper also includes appropriate performance analysis of said method in comparison with existing methods.
Keywords : component; Virtual Reality; Gestures; Arduino; SVM Classifier;Sensors; Augmented Reality; Mixed Reality.