HANDWRITTEN CHARACTER RECOGNITION USING METHOD FILTERS

Abstract

Handwritten character recognition is an emerging and a very challenging field of research as the handwritings vary from person to person. In this paper we have focused on some of the existing methodologies of character recognition and come up with some new methodologies. A system which encompasses different character recognition methods as filters is proposed in this paper. The methods are prioritized based on their result efficiencies and applied on the input. As we pass through the process, the number of possible results in the solution set keeps decreasing steeply. Using a combination of methods as a filter for recognition yields more accurate results than using a single method and also decreases the space and time complexity of the algorithm. Finally, further scope of development of this model is discussed.