Predictive Quality Management in Nepalese In-air Handwritten Kannada Numeral Recognition Using LMC

Abstract

Traditional interfacing or interaction with computers using keyboards, mouse, etc., may become obsolete after few years due to the rapid development of human computer interactions (HCI) technology. From the last two decades, in-air hand written text recognition is one of the newest and latest way to recognize the character or text written on free air. Because of this, many researches have engaged in the design of recogntion modules for characters or texts written on air for non-Indic characters and Indic characters. In this paper, explanation of the work carried out to create in-air handwritten dataset of Kannada numerals using LMC and then the design of recognition system for the classification of Kannada numerals is given.  The collected data is subjected to pre-processing and feature extraction. The first-derivative features are extracted from the preprocessed data and k-NN is used as the classifier. The simulation experiments has given the  average recognition accuracy of 96.33%.