期刊名称：International Journal of Computer Science and Information Technologies
电子版ISSN：0975-9646
出版年度：2016
卷号：7
期号：1
页码：208-210
出版社：TechScience Publications
摘要：Challenges in handwritten characters recognition isdue to the variation and distortion of handwritten characters,since different people use different style and way of drawthe same shape of the characters. This paper demonstratesthe nature of handwritten characters, conversion ofhandwritten data into electronic data, and the neuralnetwork approach to make machine capable of recognizinghand written characters. This motivates the use of GeneticAlgorithms for the problem. In order to prove this, we madea pool of images of characters. We converted them tographs. The graph of every character was intermixed togenerate new or unique styles intermediate between thestyles of parent character. Character recognition involvedthe matching of the graph generated from the unknowncharacter image with the graphs generated by mixing.
