29-12-2010, 03:03 PM
[attachment=7744]
Submitted By:
A.K.M Bellal Hossain
Lecturer Department of Computer Science
A.K.M Bellal Hossain
Lecturer Department of Computer Science
Abstract:
This paper present a recognition system for hand written character using artificial neural networks. We use Kohonen Self Organization Map for pattern classification and recognition which employs unsupervised learning algorithm. The out come of this paper shows very encouraging performance for hand written character recognition.
Introduction:
Character recognition is the process to classify the input character according to the predefine character class. With increasing the interest of computer applications, modern society needs the handwritten text into computer readable form. This research is a simple approach to implement that dream as the initial step to convert the handwritten text into computer readable form. Some research for hand written characters are already done by researchers with artificial neural networks. In this paper we use Kohonen Self Organization map. A net work, by its self organizing properties, is able to infer relationships and learn more as more inputs are presented to it. One advantage to this scheme is that we can expect the system the change with changing condition and inputs. The system consistently learns. Moreover the recognition ratio is excellent in the proposed system. According to the nature of the pattern to be recognized, recognition may be divided into two major types: the recognition of concrete items and recognition of abstract items. . However in many applications, it should be more beneficial if the network is trained form its own classification of the training data. To do this two basic assumptions about the network are made. The first is that class membership is broadly define as input patterns that share common features, the other is that the network will be able to identify common features across the range of input patterns. Kohonen’s self-organizing map is one such network that works upon this assumption, and use unsupervised learning to modify the internal state of the networks to model the features found in the training data.
