International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 68 - Number 13 |
Year of Publication: 2013 |
Authors: Pijush Chakraborty, Paramita Sarkar |
10.5120/11637-7117 |
Pijush Chakraborty, Paramita Sarkar . An Approach to Handwriting Recognition using Back-Propagation Neural Network. International Journal of Computer Applications. 68, 13 ( April 2013), 6-12. DOI=10.5120/11637-7117
Handwriting Recognition is an important topic in Computer Science. In this paper a procedure is presented to identify characters that are drawn using stylus with the help of Back-Propagation Neural Network. The paper describes the entire procedure from feature extraction to the training of the network. Feature extraction includes pixel grabbing, finding character bounds and down-sampling of the image to 7*5 pixel size. The neural network is trained by supervised learning which includes error calculation using back-propagation approach. After the training is completed the network is ready to identify characters that are drawn in the computer. This same approach given in this paper can be extended to recognize handwritten documents and also for recognizing multilingual characters.