Author : Prajakta Pandurang Shrimandilkar and R.B. Sonawane

In the office automation area, many handwritten documents are digitized for ease of backup and transmission. The demand for digital writing instruments is thus expected to grow rapidly in coming years. There are many different types of systems already available on the market. Recently, we have developed a prototype of a MEMS (Micro-Electro-Mechanical Systems) based digital writing instrument which makes use of a micro inertial measurement unit (μIMU) constructed from MEMS accelerometers and gyroscopes for real-time capture of handwriting. We have proposed an accelerometer digital pen for handwritten digit & gesture recognition. In this pen type device consist of a triaxial accelerometer, a microcontroller, and RF wireless module. The acceleration signals measured by this transmitter are transmitted to the computer by using wireless module. After measurement of this acceleration signals we can recognize it by using trajectory recognition algorithm. In the recognition algorithm basically we are using acceleration acquisition, signal preprocessing, feature generation, feature selection and feature extraction. In this project we are using the KBCS (Kernel-based class separability) for selecting the significant features from the original features. After selection of the feature it can reduce by using the LDA (Linear Discriminant Analysis) because it gives the better recognition performance

Keywords: RF wireless module, Microcontroller, Accelerometer, KBCS (Kernel- based class separability) for the feature selection, LDA (Linear Discriminant Analysis) for the feature reduction

Article published in International Journal of Current  Engineering  and Technology, Vol.4,No.4 (Aug- 2014)