Asian Journal of Engineering and Technology Innovation

Survey of An Accelerometer-Based Digital Pen for Handwritten Digit and Gesture Recognition

Sachin Ashok Waghmare , Shilpa.P. Kodgire

Marathwada Institute of Technology, Aurangabad, India

Abstract

In this paper we proposed an Accelerometer based on Micro Electro Mechanical Systems. The (MEMS) which is used in Digital pan for handwritten Digit and Gesture recognition. The (MEMS) is composed of a accelerometer, microcontroller and wireless module RF. user can hold this pen and write digit in air at normal speed. During drawing or writing movements the inertial sensor signal generated for the movements are transmitted to a computer via the wireless module. After measurement of this acceleration signal we can recognize it by using trajectory recognition algorithm. In this trajectory recognition algorithm we are using the acceleration acquisition, signal preprocessing, feature generation, feature selection, and feature extration. The algorithm is capable of translating time-series acceleration signal into a feature vector.The trajectory recognition algorithm first extract the time-and frequency-domain feature from the acceleration signal and,then further identifies the important feature by KBCS (Kernal- based class separability) for selecting significant feature and then for reducing the dimension of feature used as LDA(linear discriminant analysis). Finally with the help of probabilistic neural network we can recognition the digit.

Keywords:

Accelerometer, gesture ,handwritten recognition, probabilistic neural network(PNN), linear discriminant analysis (LDA),RF wireless module, Microcontroller