Developing a Real-time Data Acquisition System to Prepare a 3D Accelerometer-based Air-writing Data Set of Bengali Characters
Mohammed Abdul Kader, Muhammad Ahsan Ullah, Md. Saiful Islam · 2021
Air-writing is a method of converting hand gestures in 3-dimensional space into visual linguistic characters. Many data sets are available for motion sensor-based air-writing of English characters. However, there is no motion sensor-based air-writing data set is found for Bengali characters. This paper represents a real-time data acquisition system which is developed to prepare a motion sensor-based air-writing data set of Bengali characters. A portable device is developed to capture the motion of the hand during air-writing. Also, a MATLAB application is developed to collect the data from the portable device via Bluetooth. The application has the option to visualize, pre-process and analyze the collected air-writing data of a character before including it in the data set. Analysis is also done to justify the validity of data captured by the proposed system. This data acquisition system can be used to capture any type of hand gesture for preparing data sets or real-time gesture recognition.