Machine-Written Character Recognition Using A Supervised Machine Learning Approach

Abstract

Machine-Written character recognition is one among the abilities of computers to recognize input characters from external sources such as e-forms, e-letters, and e-documents. The field of character recognition can make appropriate ways to regenerate such a text format from official electronic documents, providing textual information for different enterprise class applications such as G2G, B2B, and B2C. Machine learning algorithms expose a great ability to intelligible interpret the input machine-written characters to digitized characters which could be stored in a particular le for further processing purposes. The objective of this research project is to implement an offline machine-written character recognition system for lower case English characters. I show that a machine-written character recognition system could be implemented using LVQ algorithm as a supervised classification strategy. Several experiments have performed on synthetic data to validate the speed and accuracy of the proposed method.